KR101534159B1 - Method and System on Patent Information Association Analysis System for Social Network Analysis Result - Google Patents

Abstract

The present invention relates to a fusion analysis method and system for providing SNA (Social Network Analysis) information on patent information.

According to the present invention, there is an advantage that fusion analysis results for SNA between patent classification symbols, between co-inventors, co-filed human, common patent, and core technology keywords are provided.

Patent information, SNA (Social Network Analysis), convergence, patent analysis

Description

TECHNICAL FIELD The present invention relates to a patent information fusion analysis method and system for generating a social network analysis result,

The present invention relates to a fusion analysis method and system for providing SNA (Social Network Analysis) information on patent information.

Patent information has the characteristics of technical information, right information and management information, and its importance is increasing in global international competition. It is possible to know technical development trends as technical information, technical ideas applied to individual patents, understand the scope of rights of individual patents as rights information, the degree of domesticization and domesticization, and grasp technology development trends of competitors as management information.

There are many ways to obtain patent information, but 1) it uses the patent information system provided by each country's patent office, 2) uses patent information system developed by private enterprise, and 3) And the use of a patent information system that is tailored. 1) is a patent information system provided by the Korea Intellectual Property Office (www.kipo.go.kr) and the Korea Intellectual Property Office (www.kipris.or.kr), and 2) .com, www.patolis.co.jp in Japan, www.wips.co.kr in Korea, and www.wisdomain.com.

Companies in each country, such as Thomson Scientific, the operating company of www.delphion.com, set up a patent information database and provide search results in a variety of ways to the search terms entered through the patent search engine. In addition, various kinds of analysis software based on these patent information have been developed, and software that is popular with the brand name AUREKA developed by the operator of www.micropatent.com (merged with Thompson Scientific) is famous. It is known that large-scale advanced technology companies in the world have built and operated a patent search and patent management system in the inside, but it is difficult to access the patent search and patent management system of the companies from outside.

In order to construct such a patent search and patent management system, it is common to spend more than several billions of dollars, and most domestic and foreign medium and small-sized venture companies are not building their own patent search and patent management system. These mid- to small-sized venture companies have a patent-dedicated organization, and they are collecting patent information for management and own company. However, companies that are small in size or not conscious of patents often do not have such a patented organization. These companies have been submitting to patent law offices or law firms to manage their own patents.

These patents are collecting and managing patent information on a platform that they provide by using online and offline patent information service providers such as www.delphion.com. Most of the patents filed under the name of the company are made by the inventors belonging to the company, and the inventor becomes a global trend to collect patent information at home and abroad related to his invention. The reason is that despite the existence of the patent publication period, more than 70% of the new technology information in the world is disclosed in the form of patent information, and patent information is more commercially available than the paper information. However, many of these inventors are not familiar with patent searches, or even those who are familiar with the search are accessing patent information service providers such as www.deljphion.com to obtain patent information.

Generally, it is common for the inventors to continuously study for a certain technology field for several years to several decades, so that it is common that the inventive technology field does not greatly change. Therefore, the patent technical information of domestic or foreign countries concerning the field directly or indirectly related to the technical field invented by the inventor will be useful information to the inventor. Therefore, it has been demanded not only to individual companies but also to the inventor's unit, and to supply differentiated patent information service system for only the specific inventor by optimizing to the inventor's unit. Such a patent information service system is developed It will greatly improve the accessibility of patent information to medium and small-sized venture companies that can not be maintained, and it will become a new platform for in-house patent information distribution for companies that can develop and maintain themselves.

In addition, companies or inventors automatically analyze the patents they manage and analyze information about their competitors and competitive technologies, monitoring information about competitors and competitive technologies, etc., If provided, it would be convenient for the enterprise or inventor.

The present invention is directed to a method and system for analyzing a fusion between a patent classification symbol, a joint applicant, a co-inventor, a citation document, and a key technology keyword including a patent classification symbol information on a patent document information.

According to an aspect of the present invention, there is provided a system for providing patent information to a user's computer, the method comprising: receiving selection information on a set of patent documents; Specifying a set of patent documents to be processed based on the selection information; Further comprising the step of generating processed patent information by performing at least one predetermined process on the specified set of patent documents to be processed, the method comprising: specifying the set of patent documents to be processed; and performing the predetermined calculation At least one of the patent classification symbols is used in at least one of the steps of generating patented patent information, and the utilization of the patent classification symbol is performed at any one of the steps of obtaining the selection information and generating the processed patent information When the first patent classification symbol is given, if the first patent classification symbol corresponds to a predetermined level on the patent classification scheme to which the first patent classification symbol belongs, And the patent information corresponding to the symbol is included in the patent information, The symbol may be any one or more of the International Patent Classification (IPC), the US Patent Class (USPC), the Japanese File Index (FI), the Japanese F-term (FT) and the European Patent Classification (ECLA) ≪ / RTI >

The predetermined level is a patent classification symbol of a dot level and the patent classification symbol of the dot level includes a dot in a title information which is a description corresponding to a patent classification symbol in the patent classification system, Is preferably included.

If the patent classification symbol is the international patent classification (IPC), the dot classification of the dot level includes a dot in the title information which is a description corresponding to the patent classification symbol among the patent classification symbols below the subgroup In the case of the United States Patent Class (USPC), a dot is included in title information, which is a description corresponding to a patent classification symbol among subclasses and the following patent classification symbols, and a Japanese file index In the case of the Japanese F-term (FT), the title code is a title code corresponding to the patent classification symbol. (Dot) is included in the title information, which is a description corresponding to the patent classification symbol among the patent classification symbols, and in the case of the European Patent Classification (ECLA) It is preferable that a dot is included in the title information which is a description corresponding to the patent classification symbol among the patent classification symbols below the group.

The information processing to include the patent information corresponding to the sub-patent classification symbol of the first patent classification symbol is performed when the first patent classification symbol is a patent classification symbol of a dot level, It is preferable that at least one of all the sub-patent classification symbols and the set of patent documents corresponding to all the sub-patent classification symbols of the first patent classification symbol is subjected to information processing.

The method for receiving the selection information includes: 1) a method of receiving input from a user computer through a search formula; and 2) information constituting a bibliography included in the predetermined patent document presented on the user's computer from the user's computer A method for receiving selection information on a set of specific patent documents selected from among a set of at least one patent document previously generated by the user and 4) a method for receiving selection information for at least one It is preferable that at least one of the above-mentioned methods is used.

The step of specifying the set of patent documents to be processed includes at least one first patent classification symbol in the course of the implementation of the methods 1) to 4) when any one of the methods 1) to 4) And when the first patent classification symbol is a patent classification code of a dot level, a set of patent documents corresponding to all the lower patent classification symbols of the first patent classification symbol of the dot level is specified as a set of patent documents to be processed .

A method in which a set of patent documents corresponding to all the sub-patent classification symbols of the first patent classification symbol at the dot level is specified as a set of patent documents to be processed is characterized in that all the sub-patent classification , And a set of patent documents to be processed is specified by a search formula or a quality expression including the first patent classification symbol and all the sub-patent classification symbols.

The individual patent constituting the patent document set includes at least one patent having weighted according to at least one predetermined weighting policy, and the processed patent information is generated by reflecting the weight given to the individual patent .

Wherein the weighting policy is calculated to include at least one of 1) predetermined counting information related to the individual patent, and 2) predetermined calculation value information, and the counting value to be a target of the preset counting information is 1) The number of families in at least one country of the first country and the second country, 2) the number of claims, and 3) the number of persons cited. The predetermined calculation value information includes 1) At least one evaluation score for at least one evaluation score for the applicant of the patent; 3) at least one predetermined calculation value for the patent classification symbol included in the patent of the inventor; 4) At least one predetermined calculation value for the patent classification symbol included in the applicant's financial information, That the specified calculated value and 6), which is calculated to include any one or more of the calculated value at least one pre-set for the patent classification codes themselves contained in the patent are preferred.

When the processed patent information is provided to the user's computer, processing patent information corresponding to the patent classification symbol information is provided, and processing of at least one or more sub-patent classification symbols of the patent classification symbol When the patent information is requested, the processed patent information preferably provides the processed patent information corresponding to at least one or more sub-patent classification symbols of the patent classification symbol on the patent classification scheme including the patent classification symbol .

According to an aspect of the present invention, there is provided a system for providing patent information to a user's computer, the method comprising: receiving selection information on a set of patent documents; Specifying a set of patent documents to be processed based on the selection information; And generating processed patent information by performing at least one predetermined process on the specified set of patent documents to be processed. The method for obtaining the selection information includes: 1) A method for receiving selection information on at least one selected from information constituting a bibliography included in a predetermined patent document presented in the user's computer from a user's computer, A method for obtaining selection information for a specific set of patent documents selected from at least one set of patent documents made by the system, and 4) a method for receiving at least one predetermined condition from the system is used , And the processed patent information is stored in the specified processing table 1) time, 2) a patent classification symbol, 3) an applicant, 4) an inventor, and 4) an inventor, , 5) an agent, 6) a country, 7) a keyword, and 8) a fusion pattern, and the multidimensional analysis information includes at least one of at least one individual dimension And a processing result value for at least one or more analysis processing algorithms, the at least one predetermined counting value, the at least one predetermined calculation expression, the at least one query result value, Provide a method of providing.

When generating the analysis information, the analysis information may include information about (1) a method of processing all patent classes, (2) a method of processing only the main patent classification, and A method of treating all patent classes with equivalent weights, and b) a method of processing main patent classification and patent patent It is preferable that the analysis information is generated by applying at least one selected from among methods of processing the sub-patent classification with different weights.

The analysis processing is a fusion analysis processing between patent classification symbols. The fusion analysis processing between the patent classification symbols includes a patent document including two or more patent classification symbols on the same patent classification system among the patent documents included in the processing target patent set It is desirable to perform at least one of association analysis, frequent pattern analysis, correlation analysis, and clustering among the two or more patent classes included in the patent documents.

When two or more patent classification symbols are included in one patent document in the same patent classification system in the fusion analysis processing between the patent classification symbols, the two or more patent classification symbols may be the highest patent The fusion analysis process includes a top patent classification symbol up to a predetermined level of the two or more patent classification symbols for each of the two or more patent classification symbols if the classification symbol is not a classification symbol or a classification symbol of a predetermined level And the like.

The convergence analysis processing may be performed for individual patent documents constituting the set of patent documents to be processed by 1) fusion analysis processing between at least two technical keywords generated based on text information extracted from at least one field of one patent document , 2) the analysis of the convergence between the co-inventors on the patent document with the co-inventor, 3) the analysis of the convergence between the co-applicants on the patent document with the co-applicant, and 4) the citation included in the citation information of one patent document And the fusion analysis processing between the patent documents. The fusion analysis processing is an association between the two or more technical keywords, the co-inventor, the co-applicant, and the joint patent documents included in the patent documents, Analysis, frequent pattern analysis, correlation analysis, and clustering. That the will is preferred.

And visualizing the result of the fusion analysis processing into a network diagram including at least two nodes, wherein the visualization processing includes a step of displaying information on the node and information on a line connecting the node and the node And more preferably at least one.

Wherein the information about the line connecting the node and the node further includes numerical information and the selection information for the numerical value or the line is received from the user computer; Extracting, from the patent information database including the set of patent documents to be processed, the found patent document information corresponding to the numerical information; And transmitting the extracted patent document information to the user computer.

Receiving change information about the set of documents to be processed from the user computer for the document set to be processed; And a step of specifying a set of documents to be processed based on the received change information, wherein the change information is obtained by providing an interface for obtaining change information to the user computer, 2) a time interval selection interface; 3) a country selection interface; 4) a patent classification symbol selection interface; 5) a citation specialized interface; and 6) ) Family information selection interface, and 7) a post-search selection interface, wherein the time type interface includes an interface for providing a selection of at least one of a filing date, a disclosure date, and a registration date, For each of the start and end sections / Month / week / day, wherein the country selection interface comprises an interface providing at least one selection for at least two countries, the patent classification selection interface comprising: An interface providing a selection of at least one of a class classification symbol selection interface, a patent classification symbol level selection interface, a main patent classification symbol, and a subpublic classification selection interface, wherein the cited patent interface includes a front citation and a rear citation An interface for providing a selection of one or more of a citation direction interface for any one or more selections, a citation depth interface for selection of a citation depth, and an interface for selection of at least one of direct citation and indirect citation, Wherein the selection interface includes an interface for providing a selection of at least one of whether or not to include family information and a selection of a country range, wherein the post-search selection interface is configured to select, as a search target, A search result corresponding to the search word is obtained, and at least one selection information is obtained from the search result, and only a set of patent documents related to the selection information It is preferable to include an interface for processing the document set to be processed.

The analysis processing is a categorization analysis processing of a set unit that takes at least two patent documents as one analysis unit, and the citation analysis processing of the set unit is performed as follows: 1) analysis processing of a patent document set subject to a front citation; 2) Subject analysis processing, and citation generation self patent document set subject analysis processing, and the analysis processing of 1) to 3) above is performed in a set of patent documents to be subjected to the analysis processing of 1) to 3) And the at least one predetermined multidimensional analysis information is generated, and the specification of the set of patent documents for forward citation for the patent document set subject analysis processing for forward citation is at least one depth for at least one person included in the set of patent documents to be processed And a set of documents consisting of citations.

The specification of the set of patent documents for the backward citation for the patent document set subject analysis is specified in such a way that it includes a set of documents to be filed later than at least one person citing the set of patent documents to be processed.

Citation generation for self-patent document set subject analysis The specification of the self-patent document set for analysis includes a set of patent documents consisting of patent documents cited at least once among the patent documents constituting the set of patent documents to be processed It is preferable that the above-mentioned method is specified by a method.

The citation analysis process of the set unit includes a method of further including a set of indirectly quoted documents for the set of patent documents at the time of specifying the set of patent documents for the front cite, Wherein the set of indirectly quoted documents includes at least one of the set of patent documents, and the set of indirectly quoted documents includes at least one set of patent documents, Wherein the indirect posterior citation document set comprises a patent document citing the patent document cited by the patent document to be processed at the same time as the filing date is ahead of the filing date of the patent document, The filing date is later than the patent document to be processed, It would patent documents citing target patent document consisting of patent documents that are cited at the same time is preferred.

Wherein the multidimensional analysis information is transmitted to the user computer in a tabular form including numerical information, and the selection information on the numeric value is received from the user computer; Extracting, from the patent information database including the set of patent documents to be processed, the found patent document information corresponding to the numerical information; And transmitting the extracted patent document information to the user computer.

At least one or more of the patent classification symbols may be utilized in at least one of the steps of specifying the set of patent documents to be processed and generating the processed patent information by performing the predetermined calculation, Wherein the use of the patent classification symbol is performed when the first patent classification symbol is given at least one of the step of obtaining the selection information and the step of generating the processed patent information, Wherein the information processing unit includes information about the patent information corresponding to the sub-patent classification code of the first patent classification code when the first patent classification code corresponds to a predetermined level on the system, IPC), United States Patent Classification (USPC), Japan File Index (FI), Japan F-term (FT), European Patent (ECLA). ≪ / RTI >

A method for receiving selection information on at least one selected from information constituting a bibliography included in the predetermined patent document presented on the user computer from the user computer includes receiving at least one applicant from the user computer Providing at least one of applicant information, patent classification information including at least one patent classification symbol, inventor information including at least one inventor, and proxy information including at least one agent, to the user computer; Obtaining selection information on at least one of the applicant, the patent classification code, the inventor information, and the proxy information from the user computer; And processing the set of patent documents identified by the selection information as a set of patent documents to be processed to generate a predetermined multidimensional analysis result and providing the result to the user computer.

The system generates at least one intermediate multidimensional analysis result table for easily and quickly generating the multidimensional analysis results for each applicant, the patent classification symbol, the inventor or the agent, A multidimensional analysis list or menu for at least one applicant, at least one multidimensional analysis list or menu for at least one patent classification, a multidimensional analysis list or menu for each at least one inventor, Providing one or more of the menus; Obtaining selection information for the multidimensional analysis list or menu from the user computer; And if the analysis list obtained from the selection information is available from the intermediate multidimensional analysis result table, the analysis result obtaining unit obtains the selection information by referring to the intermediate multidimensional analysis result table and provides the analysis result to the user computer The method further comprising:

According to an aspect of the present invention, there is provided a system for providing patent information to a user's computer, the method comprising: receiving at least one patent document information disclosed in a first country; If the first station filing date and the date of receiving the selection information are more than one year, the patent document having received the selection information includes 1) PCT application information and 2) patent document of the first country in the second country And 3) a current state information of a patent document of a second country corresponding to the second station family information; If there is no PCT application information for the patent document, and the second station does not have the family information, even if there is the PCT application information, A second case in which there is no second country family information, and a second case in which the current state information of the second country patent document is withdrawn even if there is the second country family information, 3; < / RTI > And when the patent document of the first country corresponds to any one of the first case, the second case and the third case, when the patent document information of the first country is displayed, The display information corresponding to one of the first case, the second case and the third case can be displayed in a predetermined vicinity of the patent document information of the first country, And transmitting the document information and the predetermined display information, wherein the second station is at least one or more than one.

The effects of the present invention are as follows.

Patent classification symbols and technology keyword information are processed, and the fusion tendency between the patent classification symbols and the technology keywords can be grasped, and the fusion property can be utilized in the patent analysis of the present invention.

 Patent information can be processed independently of the size and type of document sets by processing various analyzes including citation analysis in units of a patent document set.

By analyzing joint applicants, co-inventors, co-leading patents together, patent classification codes together, and technology keywords together, network and technology between networks and patent classification symbols between applicants and inventors, The network between the keywords can be grasped, and analysis results for each of them can be generated.

Premise information

The format of the published patent document and the information it contains

The Patent Office of each country issues a public notice about patent application or patent registration with certain conditions. The patent information that is disclosed differs from country to country but includes information about the bibliography and information about the patent text.

The information on the bibliography includes information on the patent document issuing country, the name of the invention, information on the applicant (including assignee, same as follows) / patentee (including assignee, hereafter), information on the inventor, Information about various dates (such as IPC, USPC, FI, FT, ECLA, etc.), information on various dates such as the filing date, and application numbers are indispensably included. In addition, if there are abstract information, representative information, information about the representative figure, reference information (reference information about the prior art initiated by the applicant, information examined or examined by the examiner, information about the patent classification symbol, etc.) Information about the priority, information about the examiner, and the like may optionally exist. The contents of the bibliography may further include information relating to the abstract, the representative claim, or the information related to claim 1 of the claims.

The text contains information on the technical content of the patent. The text of a typical patent document contains information on the name of the invention, information on the composition of the invention, and information on the scope of the patent claim. In addition to these, the effects of the invention, industrial applicability, Description of the drawings, and conventional techniques. In addition, information relating to drawings is essentially or optionally included depending on the inventor's choice or the type of patent technology.

Patent Classification Symbol Information and Patent Classification Sign System exists

In addition, each patent document contains at least one patent classification symbol on at least one of the patent classification schemes. The IPC (International Patent Classification) symbols are assigned to documents issued by countries such as Korea, the United States, Japan, and Europe (EPO). Each country has its own unique classification system (USPC Or UPC, for the Japanese Patent Office F-

Term / FT, FI (File Index), and European Patent Office (ECLA)), the patent classification code of the issuing country of the patent document is given.

The patent classification schemes have a 1: 1 correspondence in each classification symbol that describes the content of the corresponding patent classification symbol "title". On the other hand, in a case where the depth is less than a predetermined depth on each patent classification system, dots are given to the title information of each patent classification symbol. From the following example, the presence of dots providing information on the hierarchical nature of the patent classification system, the presence of title information, and the relative position of the classification system in front of the title information is shown.

Section: H Electricity

Class: H01 Basic electrical element

Subclass: H01F Magnets

Main group: H01F 1/00 Magnet or magnetic body characterized by magnetic material

1-dot subgroups: 1/01 * Inorganic materials

2-dot subgroups: 1/03 ** characterized by coercivity

3-dot subgroups: 1/032 *** of hard magnetic material

4-dot subgroup: · 1/04 **** Metal or alloy

5-dot subgroup: 1/047 ***** Alloy characterized by composition

6-dot subgroup: 1/053 ****** Containing rare earth metals

Taking the IPC as an example, the patent classification symbols corresponding to all the subgroups have a structure of "subclass" + "number / number", and titles corresponding to the patent classification codes are combined with dots (Usually, the numbers / 00 are the symbols corresponding to the main group.) The smaller the number of dots combined in a title, the more the title is related to its surroundings In comparison, the technical classification is a relatively higher concept, and the higher the number of dots, the more the technical classification of the sub / detailed concept is as compared with the surrounding area.

As in the above example, the IPC has a multi-level hierarchical structure (tree structure). On the other hand, USPC, FT and ECLA also have a multi-level hierarchical structure. The multi-level hierarchical structure of each patent classification symbol may be DB.

Introduction of the configuration of the patent information system (1) of the present invention

First, the structure of the patent information system 1 of the present invention will be briefly described.

1 and Fig. 13.

The patent information system 1 of the present invention is connected to a user computer 300 through a wired / wireless network 200. At this time, the user computer 300 may be a computer based on a person such as a PC, a notebook computer, a wired / wireless communication terminal, or a game machine, a server of a third server, organization or institution other than the present patent information system 1, A program module, or the like, connected to the patent information system 1. When the patent information system 1 performs a web service, the patent information system 1 must include a web service support module.

The patent information system 1 includes a DB unit 20 related to various data such as patent data, a preprocessing module 30 for performing various preprocessing, a patent information processing basic module 40 for processing and analyzing patent information, (A member information processing module 501, a multilingual processing module 502, a translation module 503, and the like are mounted) for carrying out various kinds of support such as membership management and policy management, The patent intelligence module 60 is referred to as a patent analysis intelligence module or a brief analysis module since the patent intelligence module generates advanced patent information mainly through patent analysis. In the present specification, the term " The same shall apply hereinafter) and a hierarchically integrated patent information record that generates a set of patent information services in the applicant or inventor or agent unit There are bus systems.

The patent information system 1 includes six components.

The first component is various preprocessing modules, which are modules that perform various preprocessing for each purpose for patent documents obtained.

The second component is the patent information processing basic module 40. These functions include 1) search function, 2) analysis function, 3) monitoring function, 4) analysis function, 5) patent document collection function, 6) , And a module for performing a reporting function.

The third component is the patent analysis intelligence module 60, which is 1) a module for generating various advanced patent analysis information including citation analysis, comparative analysis and the like.

The fourth component is a module for generating the patent information system (1) by the applicant, the inventor, the agent, and the patent classification by the patent information service system generation module.

The fifth component is that the support module 50 allows the user to process 1) member information processing, 2) multilingual processing, 3) translation processing, 4) web service processing and other aspects of the invention can be serviced in various network 200 environments It is the component responsible for the support.

The sixth component is the various DBs.

1) Patent document The mast DB 202,

2) Patent classification symbol mast DB 203,

3) various mast DBs such as the subject mast DB 204,

4) Various DB related to analysis,

5) Various support DBs (member DB 206-2, menu DB 206-3, policy DB (representative phrase extraction DB 206-1-1, weighting policy DB 206-1-2, Patent index DB 206-1-3, and analysis quality DB 206-1-4)),

6) Various secondary processing DBs (text mining DB 207-1, representative word DB 207-2,

7) multiple patent classification symbol relation DB 207-3) and the like.

It should be understood that the above-mentioned six components are typical elements, and other configurations described in the present invention are included in the system 1 of the present invention, and those skilled in the art will understand the functions and modules (firewall, Etc.) are not mentioned separately.

The DB unit 20 is provided with a DBMS 201 for managing DBs, a patent document mast DB 202 which is a DB storing integrated patent documents, a patent classification code mast DB 203), a subject mast DB 204 storing information on applicants, inventors, companies, etc., a multidimensional analysis calculation execution result table DB 205 storing multidimensional analysis results, information on members and various policies and options A secondary DB part 207 storing the result of processing various kinds of data, and the like.

In the menu DB, there is a notation for each menu for each country or each language. The patent information system 1 of the present invention reads the locale information in the browser of the connecting user and stores the locale information in each country / language Extracts a star menu from the menu DB, creates a screen with the menu of the extracted language, and transmits the screen to the user.

The pre-

Various preprocessing modules of the present invention will be described. 2) a counting preprocessing module 3100; 3) a statistical preprocessing module 3200 according to a patent classification symbol; 4) a weight preprocessing module 3300 5) citation information preprocessing module 3400, 6) patent classification symbol preprocessing module 301-3-1 or 3500, 7) applicant representative name preprocessing module 301-4-1-1 or 3600, 8) A representative phrase extraction preprocessing module 3700, 9) a family information preprocessing module 3800, 10) a plurality of patent classification symbol relation preprocessing module 3900, 11) a holder change information preprocessing module 302, 12) Module 303 may be further included.

The original patent document processing module (301-1)

The original patent document processing module 301-1 of the present invention will be described. The original patent document processing module 301-1 is disclosed in each country (the internationally filed patent document is disclosed by WIPO). WIPO is not a country, but is included in each country for convenience of description. Process data (XML, SGML or other formats).

The original patent document processing module 301-1 obtains the acquired patent document data, performs predetermined processing, and then matches the acquired patent document data to a format that can be processed by the patent information system 1 of the present invention Generate patent document data

The predetermined processing may include 1) error filtering, 2) patterned error correction, 3) manual error correction, 4) conversion of SGML document data into XML document data, and 5) standard format conversion.

The original patent document processing module 301-1 passes the acquired patent document data to at least one or more error verification filters 301-1-1, (1). ≪ / RTI >

Patent document mast DB generation module (301-2)

The patent document mast DB generation module 301-2 of the present invention generates the patent document mast 301-2 based on the patent document data generated by the original patent document processing module 301-1 and the information processed by the various preprocessing modules of the present invention DB 202 is generated. The detailed modules included in the patent document mast DB generation module 301-2 of the present invention are listed as follows. 1) generate a source bibliographic item DB 202-1-1, a bibliographic item DB 202-1-2, a translation bibliographic item DB 202-1-3, and a representative drawing DB 202-1-4 And a bibliography DB generation module 301-2-1. 2) a document DB creation module 301-2-2 for creating a patent document DB 202-2-1, a translation patent document DB 202-2-2, and a processed patent document DB 202-2-3; . 3) a US assignee change DB 202-3-1, and a rights holder change DB creation module 301-2-3 that generates a rights setup change DB 202-3-2. 4) a family information DB generation module 301-2-4 for generating a country-specific patent status DB 202-4-1 and a country-specific family DB 202-4-2. 5) a citation information DB generation module 301-2-5 for generating the applicant citation information DB 202-5-1 and the examination citation information DB 202-5-2, and 6) the citation information DB 202- 6) for generating the processing result information.

The patent document master DB 202 includes a bibliographic database, a document database, a rights change information database, a family information DB, a citation information DB, and an administrative processing information DB.

The bibliographic database includes bibliographic information directly obtained from the patent document data, processed bibliographic information processed for the patent document data itself or the whole patent document data set, the bibliographical references and bibliographic references And includes translation bibliographic information and other information translated by the translator. The bibliography DB generation module of the present invention is responsible for generating the bibliographic database.

The document DB includes patent document data on the patent document data itself, translational document data for each language in which part or all of the patent document data is translated into various languages, data on the patent document data or translated patent document data on the language into pdf, doc , tiff, html, and so on. The patent document data, the translated patent document data and the processed patent document data by language may exist in the form of a patent document DB, a translation patent document DB, and a processed patent document DB, respectively. The document DB creation module of the present invention is responsible for creating the document DB.

The rights holder variation information DB includes an assignee variation information database for US patent documents. Information on changes in the names of various patent owners and patent applicants managed by each country's patent office, registration and erasure registration of various licensees As shown in FIG. The owner variation DB generation module of the present invention generates the owner variation information DB.

The family information DB includes the family data for each country and the status of each family patent document in each country (including the current state of the application by country, the right to obtain the rights, etc.) ) Information about the patent status of each country. The generation of the family information DB is performed by the family information DB generation module of the present invention.

The citation information DB includes an applicant's citation information DB generated based on the citation information cited by the patent applicant, a citation information DB cited in the examination process of the examiner for the specific patent application (mainly referred to as " Document) There is an examination citation information DB which is generated based on information on examination of other patent documents. The citation information DB generation module of the present invention is responsible for generating the citation information DB.

The administrative processing information DB includes data generated based on the administrative processing information generated by each country's patent office for each patent document. The administrative processing information DB generation module of the present invention is responsible for generating the administrative processing information DB.

Then, the bibliographic database generation module can obtain the processed bibliographic item data by obtaining information processed by various preprocessing modules, etc., on the patent document data and / or the entire patent document data, The included DB is called the bibliographic DB. The processing bibliography may include 1) various counting information, and 2) various calculation and evaluation information.

4) number of pages of specification pages, 5) number of drawings, 6) number of documents, 6) number of documents, 6) number of documents, 8) Number of references (= number of citations) / number of citations, 9) Number of examination citation / number of citations, 10) Number of patent classification codes The number of patent classification marks, 11) the number of priority claims, 12) the number of family patents in each country including its own country, 13) the total number of families, 14) 17) the total number of patent documents in the reference, and 18) the number of non-patent documents in the reference. The counting preprocessing module of the present invention processes such counting information. Details will be described in the counting preprocessing module of the present invention.

In the case of various calculation and evaluation information, for example, (1) the patent classification mark included in the patent document of the applicant included in the patent document (the patent classification mark itself, the entire subpopulation classification symbol of the patent classification mark, Such as the total number of applications, the total number of registrations, the percentage of shares, the concentration rate, the activity rate (AI), etc. for the patent classification code of the next higher level or the patent classification code of all stages up to the main group of the patent classification code) 2) a value obtained by calculating various patent indices of the patent classification mark included in the patent document of the inventor included in the patent document, and 3) various analysis indexes. For details, Various preprocessing modules and analysis modules according to the present invention will be described.

Patent classification code mast DB generation module 301-3

The preprocessing module of the present invention includes a patent classification code mast DB generation module 301-3. The DB generated by the patent classification code mast DB generation module 301-3 is as follows.

1) The original patent classification code database 203-1 storing raw data of various patent classification codes obtained from patent offices or patent source data sources of respective countries.

2) a modified patent classification symbol DB 203-2 storing a patent classification code in which various patent classification codes are modified in accordance with the use of the patent information system 1 of the present invention, 301-3-3).

3) a total upper patent classification symbol set DB 203-3 that collects and stores all upper patent classification symbols for an arbitrary patent classification symbol, which is a total upper patent classification symbol set generation module 301-3-1- 1).

 4) a sub-patent classification symbol set DB 203-4 storing information on a sub-patent classification symbol or any sub-patent classification symbol for any patent classification symbol, and this patent classification symbol set generation module 301-3- 1-2). The sub-patent classification symbol set generation module 301-3-1-2 includes a sub-patent classification symbol set generation module 301-3-1-2-1 for generating only a sub-subset of patent classification codes, And a total sub-patent classification symbol set generation module 301-3-1-2-2 for generating a classification symbol set.

5) Patent classification symbol tree table DB 203-5 storing the patent classification symbol system as a tree structure, which is generated by a patent classification symbol tree table creation module (not shown).

6) There is a total upper patent classification symbol table DB (203-6) storing the total upper patent classification symbols for any patent classification symbol by levels, which is the total upper patent classification symbol set generation module 301-3-1 -1). On the other hand, when there is an update in the patent classification symbol, it is processed by the patent classification update module 301-3-5 of the present invention, and the predetermined module related to the updated patent classification symbol performs predetermined processing.

 Hereinafter, patent classification symbols will be described collectively.

Multistage of patent classification codes

Each patent classification symbol in a multi-level hierarchical (tree) structure has a patent classification symbol. And the corresponding relationship of the title information of the patent classification symbol. A number of such examples have been presented elsewhere in this patent specification. (Search module 401 and search engine are used as synonyms in the present invention) or DBMS 201 to obtain information on all of the sub-patent classification codes of a given patent classification symbol on the hierarchical structure, Card or the like can be used. For example, to obtain all the information about sub-patent classification of H01F, all sub-patent classification codes of H01F can be obtained by querying the query or query by combining the extension / wildcard (eg?) With H01F . Similarly, "H01F1?" Can be input to obtain all information about the sub-patent classification of H01F1 / 00. However, it is not possible to enter "H01F1 / 01?" In order to obtain all information on the sub-patent classification of H01F1 / 01. The reason is that the patent classification symbols including the dot in the title information of the patent classification symbol such as IPC have the same display pattern (for example, in the case of IPC, the main group indication + / + number, in case of USPC, The number of dots, and the relative parent-child relationship / upper-level relationship are displayed while having the number of dots in the FT case. In other words, if the title information does not include a dot, it is possible to distinguish the parent-child relationship / parent-child relationship only by the notation system of the patent classification symbol, but otherwise it can not be distinguished only by the notation system of the patent classification symbol .

On the other hand, in the property that the patent classification symbol has a hierarchical structure, it is desirable that the patent document corresponding to the sub-patent classification code corresponds to the upper patent classification code. That is, if a patent classification code H01F1 / 04 is assigned to a specific patent document, this patent document also applies to H01F1 / 032, corresponds to H01F1 / 03, corresponds to H01F1 / 01, and also corresponds to H01F1 / 00 . In reverse, the patent information pertaining to H01F1 / 03 shall include not only H01F1 / 03, but also patent information related to all patent classification codes under H01F1 / 03 in the hierarchical structure of the patent classification symbol. In other words, patent information pertaining to H01F1 / 03 should include not only H01F1 / 032 and H01F1 / 04, but also patent information related to all patent classification codes under H01F1 / 03 in the patent classification system.

FIG. 14 shows an example in which the inventive idea of the present invention is not applied. As can be seen from Fig. 14, there are many documents corresponding to the sub-category A61B 3/02 than the documents corresponding to A61B 3/00.

3 ways of processing sub-patent classification

For the above reasons, it is necessary to process information on all the sub-patent classification symbols of the patent classification symbol in which the dot is included in the title information. The patent classification symbol preprocessing module 301-3-1 or 3500 included in the patent classification code mast DB generation module 301-3.

It is necessary to include patent information on the sub-patent classification in all the processes of information extraction for a given patent classification. 1) Search, 2) Statistics, 3) Analysis, 4) Monitoring, and 5) directory display. In the case where the extension (wildcard) can not be used, in order to include information on all the sub-patent classification symbols for a given patent classification, essentially one of the following processes is required.

First, the patent classification system (tree structure) is searched for a given patent classification symbol in a manner such as a depth first search to obtain information about all sub-patent classification symbols. This information retrieval process can occur every time a query / retrieval occurs. On the other hand, all the sub-patent classification symbols for the patent classification symbols including dots in all patent classification symbols or title information are obtained and stored for a given patent classification scheme (tree structure), and all sub-patents stored in the query / It may be possible to obtain information on the classification symbol. Given the data in a tree structure, it is essential for each node constituting this tree structure to find a node directly below the node and / or all its descendants in a manner such as depth-first search, Therefore, a description thereof will be omitted. In addition, storing the information of the lower node and / or all of the lower nodes obtained in correspondence with the specific node is also a basic one, so that the explanation thereof is also omitted.

Second, for all patent classification codes, information about the patent classification code corresponding to the upper part of the patent classification code is stored in a stepwise manner with reference to the patent classification system (tree structure). Then, when there is a given patent classification code, It is possible to find a specific step that appears first, and to obtain information about all patent classification symbols for which the patent classification mark is indicated at the specific step.

Third, a given patent classification scheme (tree structure) can be modified to generate only the representation of the patent classification symbol, generating the modified patent classification symbol, and perform the first and second methods for the modified patent classification symbol . On the other hand, by appropriately selecting a method of transforming referring to the patent classification system, it is possible to modify the patent classification notation so as to respond to a range search / range query.

Illustrative explanation through H04B 7/00

In the present invention, the tree structure between the patent classification symbols is described by taking the patent classification code corresponding to IPC 7th edition reference main group H04B 7/00 (title is a wireless transmission system) and its lower part as an embodiment, The method and description are not limited to the present embodiment, but may be applied to the same IPC area or the entire IPC domain, and may be applied to other patent classification schemes (for example, For example, USPC, FI, FT or ECLA) are applied in the same or equivalent manner.

The patent classification codes and titles corresponding to the main group H04B 7/00 and its lower part shall have the following composition on the basis of IPC 7 edition.

H04B 7/00 Radio transmission systems, ie using radiated electromagnetic fields

Control of transmission H04B 7/005. Equalization

H04B 7/01 Reduction of phase shift

H04B 7/015. Reduction of echo effect

Diversity systems H04B 7/02.

H04B 7/04 .. using a plurality of discrete aerials spaced apart

H04B 7/06 ... in the sending station

H04B 7/08 ... in recipient station

H04B 7/10 .. using a single aerial system characterized by polarization or directional characteristics

H04B 7/12 .. Frequency Diversity System

H04B 7/14 Relay system

H04B 7/145 .. Manual relay system

H04B 7/15 .. Active relay system

H04B 7/155 ... ground-based stations

Using angle modulation H04B 7/165 ....

H04B 7/17 .... using pulse modulation

H04B 7/185 ... space-based or airborne stations

H04B 7/19 .... earth synchronization stations

H04B 7/195 .... Asynchronous station

H04B 7/204 ... multiple access

H04B 7/208 .... frequency division multiple access

H04B 7/212 .... Time division multiple access

H04B 7/216 .... Code division or spread-spectrum multiple access

H04B 7/22. Spawner wave propagation system

For communication between two or more points H04B 7/24.

H04B 7/26 .. that at least one point can be moved

The tree hierarchy structure can be expressed as shown in Table 1 below.

[Table 1]

Main group 1-dot subgroup 2-dot subgroup 3-dot subgroup 4-dot subgroup title H04B 7/00 Wireless transmission system H04B 7/005 Control of transmission; Equalization H04B 7/01 Reduction of phase shift H04B 7/015 Reduction of echo effect H04B 7/02 . Diversity system H04B 7/04 Using multiple independent antennae spaced apart H04B 7/06 ... in the sending station H04B 7/08 ... in the receiving station H04B 7/10 Using a single aerial system characterized by polarization or directional characteristics H04B 7/12 .. Frequency Diversity System H04B 7/14 Relay system H04B 7/145 .. Manual Relay System H04B 7/15 .. Active Relay System H04B 7/155 ... ground-based stations H04B 7/165 .... using angle modulation H04B 7/17 .... using pulse modulation H04B 7/185 ... space station or aircraft carrier H04B 7/19 .... Earth Motive Bureau H04B 7/195 .... asynchronous station H04B 7/204 ... multiple access H04B 7/208 .... frequency division multiple access H04B 7/212 .... Time Division Multiple Access H04B 7/216 Code division or spread-spectrum multiple access H04B 7/22 .Shelter Wave Propagation System H04B 7/24 For communication between two or more points H04B 7/26 ... that at least one point can be moved

At this time, referring to the patent classification system (tree structure), if all the patent classification codes at the top of the patent classification system are obtained, a total top patent classification table as shown in Table 2 below can be generated. It will be appreciated that the table of total top patent classification symbols as shown in Table 2 below can be generated not only for IPC but also for USPC, FT, ECLA, FI and the like. On the other hand, when the patent classification symbols are updated, it is preferable that the upper general patent classification symbol table as shown in Table 2 below is updated and the table is updated as well. The total top patent classification symbol table as shown in Table 2 below is obtained by obtaining the patent classification symbol from the patent classification symbol mast DB of the total top patent classification symbol table generation module of the present invention. The total upper patent classification symbol table generation module preferably generates the total upper patent classification symbol table every time the information on the patent classification symbol is updated. The total upper patent classification symbol table generation module may generate a total upper patent classification symbol table of the same type for the modified patent classification symbol DB of the present invention.

In the following Table 2, IPC_ID indicates a specific IPC ID, S indicates a section level, C indicates a class level, SC indicates a subclass, MG indicates a main group, 1 dot indicates a level having one dot in the title information, (Therefore, n dot indicates a level having n dots in the title information), self indicates a level having two dots in the title information, and 3dot indicates a level having four dots in the title information. .

[Table 2]

IPC_ID S C SC MG 1dot 2dot 3dot 4dot Self 69964 H H04 H04B H04B7 / 00 H04B7 / 00 69965 H H04 H04B H04B7 / 00 H04B7 / 005 H04B7 / 005 69966 H H04 H04B H04B7 / 00 H04B7 / 01 H04B7 / 01 69967 H H04 H04B H04B7 / 00 H04B7 / 015 H04B7 / 015 69968 H H04 H04B H04B7 / 00 H04B7 / 02 H04B7 / 02 69969 H H04 H04B H04B7 / 00 H04B7 / 02 H04B7 / 04 H04B7 / 04 69970 H H04 H04B H04B7 / 00 H04B7 / 02 H04B7 / 04 H04B7 / 06 H04B7 / 06 69971 H H04 H04B H04B7 / 00 H04B7 / 02 H04B7 / 04 H04B7 / 08 H04B7 / 08 69972 H H04 H04B H04B7 / 00 H04B7 / 02 H04B7 / 10 H04B7 / 10 69973 H H04 H04B H04B7 / 00 H04B7 / 02 H04B7 / 12 H04B7 / 12 69974 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 14 69975 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 145 H04B7 / 145 69976 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 15 69977 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 155 H04B7 / 155 69978 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 155 H04B7 / 165 H04B7 / 165 69979 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 155 H04B7 / 17 H04B7 / 17 69980 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 185 H04B7 / 185 69981 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 185 H04B7 / 19 H04B7 / 19 69982 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 185 H04B7 / 195 H04B7 / 195 69983 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 204 H04B7 / 204 69984 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 204 H04B7 / 208 H04B7 / 208 69985 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 204 H04B7 / 212 H04B7 / 212 69986 H H04 H04B H04B7 / 00 H04B7 / 14 H04B7 / 15 H04B7 / 204 H04B7 / 216 H04B7 / 216 69987 H H04 H04B H04B7 / 00 H04B7 / 22 H04B7 / 22 69988 H H04 H04B H04B7 / 00 H04B7 / 24 H04B7 / 24 69989 H H04 H04B H04B7 / 00 H04B7 / 24 H04B7 / 26 H04B7 / 26

In other words, given the IPC patent classification code H04B7 / 06, if all parent nodes (branch points in the tree structure are referred to as nodes) are found by referring to the IPC patent classification system, H, H04, H04B , H04B7 / 00, H04B7 / 02, H04B7 / 04, H04B7 / 06, and H04B7 / 06, and the data as shown in IPC_ID69970 can be generated according to the field structure.

Given the total top patent classification table as shown in Table 2 above, obtaining all the top patent classification symbols of a given patent classification symbol is accomplished by 1) searching for a given patent classification symbol in the self field, You will get the patent classification symbol for each step as you move up. The total top patent classification symbol sets of the given patent classification codes are collected and DBed together to form a database of the total top patent classification codes set of the present invention. The classification symbol is extracted and converted into a DB, which is a sub-patent classification symbol set DB of the present invention. At this time, it is possible to extract only the sub-classifications of a specific patent classification symbol using the above-mentioned upper patent classification symbol table. This information is DB, which is a subset of the patent classification set of the present invention.

When there is a total top patent classification symbol table as shown in Table 2 above, and given a specific patent classification symbol, a description will be given by way of example of H04B7 / 15 as a method of obtaining the patent classification symbol and all the lower patent classification symbols. If you find H04B7 / 15 in Self first, you will find IPC_ID69976. In step 2, if it is found at which level field that the same patent classification code is associated with itself, it is found that it is a 2-dot level field. In step 3, if you find a patent classification code with H04B7 / 15 in the 2-dot level field, you will find the patent classification code H04B7 / 15, H04B7 / 155, H04B7 / 165, H04B7 / 17, H04B7 / 185, H04B7 / 19, H04B7 / 195, H04B7 / / 204, H04B7 / 208, H04B7 / 212, and H04B7 / 216. These are all sub-patent classification codes of H04B7 / 15 itself and H04B7 / 15. It is of course possible to match all sub-patent classification codes with oneself by excluding oneself from the found result. At this time, the sub-patent classification symbol set DB can be generated from the information about the patent classification symbol and all the sub-patent classification symbols for a specific patent classification symbol. At this time, if all the sub-patent classification symbols of the patent classification symbol are stored for a specific patent classification symbol, a total sub-patent classification symbol set DB can be generated. It may also count the number of all sub-patent categories (including 11 in this case and 10, excluding oneself), including oneself. Such counting is preferably performed by the counting preprocessing module of the present invention.

If the 3-dot level field value is not null (there is a value directly below) and the 4-dot level field value is null when the 3-step level value is null (No lower value directly below) If you find the patent classification, you will get H04B7 / 155, H04B7 / 185, H04B7 / 204 respectively. At this time, if a patent classification code is stored for a specific patent classification code, the partial patent classification code set DB can be generated.

At this time, it is possible to count the number (3 in this case) of the own sub-classification codes, and this counting is preferably performed by the counting preprocessing module of the present invention. This is because of the necessity of step-by-step unfolding when the subclause is required. Step-by-step unfolding means that each level (section, class, subclass, main group, one dot subgroup, two dot subgroup, three dot subgroup, four dot subgroup, ... n dot sub (Such as search results, statistical values, calculated values, analytical values, etc.) for each step even if the display is stepwise unfolded. All values) should be taken into account for all of the specific patent classification and all sub-patent classification codes of that particular patent classification. For example, when expanding under H04B7 / 15, H04B7 / 155, The values for H04B7 / 155 and H04B7 / 155, including both H04B 7/165 and H04B 7/165, which are lower than H04B7 / 155 and H04B7 / 155, Should be displayed as a value of / 155 It means, of course, in special cases will be able to display the values for only that particular patent classification symbols (such as the user's choice).

The above table structure (referred to as the table used for finding the sub-patent classification symbol in the above three steps) can be utilized in obtaining information necessary for a search, an SQL command, and statistical analysis processing described later. (For example, a view, a materialized view, and a cube) are calculated for each patent classification symbol in advance in order to facilitate the analysis, Etc.). When calculating these calculations, roll up (roll up) is a concept that includes the value of itself as its value by including the values for its underlying objects. The above table can be utilized. For example, a roll-up operation can generate a year value by adding a quarter value to a year value, and a quarter value can be generated by adding a month value of each branch. When calculating a value for a particular patent classification (for example, the number of applications per year), a value for the patent classification can be generated by adding a value for itself and a value for its sub-patent classification. Details will be described in detail in the analysis of the present invention.

If all the IPC codes / IPC patent classification codes are processed in the same manner as above, a table having the above-described format targeting all the IPC patent classification codes can be generated, and these tables and the respective languages It is possible to generate the source IPC patent classification symbol DB corresponding to the IPC among the original patent classification symbol DB including the star title information. In the same manner, USPC, FT, FI, and ECLA can also generate the same table.

H04B7 / 00 The following is an example of an information structure that shows the child-parent relationship between patent classification symbols.

First, it is a result of a method of mapping a child patent classification symbol to parent (child -> parent), which is a parent classification code, to which the classification symbol belongs. Examples are shown in Table 3 below.

[Table 3]

Child (self) Parent (your parent) H04B 7/005 7/00 H04B 7/01 7/00 H04B 7/015 7/00 H04B 7/02 7/00 H04B 7/04 H04B 7/02 H04B 7/06 H04B 7/04 H04B 7/08 H04B 7/04 H04B 7/10 H04B 7/02 H04B 7/12 H04B 7/02 H04B 7/14 7/00 H04B 7/145 H04B 7/14 H04B 7/15 H04B 7/14 H04B 7/155 H04B 7/15 H04B 7/165 H04B 7/155 H04B 7/17 H04B 7/155 H04B 7/185 H04B 7/15 H04B 7/19 H04B 7/185 H04B 7/195 H04B 7/185 H04B 7/204 H04B 7/15 H04B 7/208 H04B 7/204 H04B 7/212 H04B 7/204 H04B 7/216 H04B 7/204 H04B 7/22 7/00 H04B 7/24 7/00 H04B 7/26 H04B 7/24

An example of the result for the method (parent - > child) having all child information on its own subordinate is shown in Table 4.

[Table 4]

Parent (self) Child (own sub child) H04B 7/00 H04B 7/005
H04B 7/01
H04B 7/015
H04B 7/02
H04B 7/14
H04B 7/22
H04B 7/24 H04B 7/005 null H04B 7/01 null H04B 7/015 null H04B 7/02 H04B 7/04
H04B 7/10
H04B 7/12 H04B 7/04 H04B 7/06
H04B 7/08 H04B 7/06 null H04B 7/08 null H04B 7/10 null H04B 7/12 null H04B 7/14 H04B 7/145
H04B 7/15 H04B 7/145 null H04B 7/15 H04B 7/155
H04B 7/185
H04B 7/204 H04B 7/155 H04B 7/165
H04B 7/17 H04B 7/165 null H04B 7/17 null H04B 7/185 H04B 7/19
H04B 7/195 H04B 7/19 null H04B 7/195 null H04B 7/204 H04B 7/208
H04B 7/212
H04B 7/216 H04B 7/208 null H04B 7/212 null H04B 7/216 null H04B 7/22 null H04B 7/24 H04B 7/26 H04B 7/26 null

Directory creation module 405,

The directory generation module 405 of the present invention provides a subclass patent classification code for each patent classification symbol and sets the count value for the patent classification symbol and the subclass patent classification code to the patent classification symbol You can mark it next to it. The counting values include 1) the number of all subclasses, 2) the number of subclasses, 3) the depth of classification, 4) the number of patent documents corresponding to all subclasses, 5) And the number of patent documents. It is preferable that 1), 2), and 5) are displayed together as much as possible. The directory generation module 405 reads one or more of the above 1) through 5) from the count value of the counting preprocessing module of the present invention, or ii) iii) the patent classification tree table, or iv) the total upper patent classification symbol table, to display the above 1) to 5).

When the above-described patent classification symbol information is arranged in a directory format, symbols and symbols such as a color or an icon are displayed in each directory, thereby improving the accessibility of the user. That is, by giving a different color or giving a predetermined icon or the like according to the case where there are many subclasses, a lot of subclasses, a deep depth, a lot of patent documents, etc., The number of subclasses, the depth of classification, the number of patent documents, etc.). In particular, if the number of all subclasses of a particular patent classification directory is large or the depth of classification is large, it indicates that there are many patent applications in the area, and that the applications are variously applied. Therefore, Is often larger.

FIG. 15 shows an embodiment of the directory creation module 405 of the present invention. As can be seen from FIG. 15, at least one patent classification symbol can be checked-boxed from a predetermined stage of the directory, and a patent classification symbol selected by the check box can be searched for by country, The patent analysis can be carried out.

The directory creation module 405 of the present invention includes a patent classification symbol utilization directory creation module 405-1 for creating a directory for a patent classification symbol corresponding to a multi-stage patent classification symbol system for each type of at least one patent classification symbol, There is a subject directory creation module 405-2 for creating a multi-level directory according to various analysis topics or other definable subject matter introduced in the specification of the present invention. An operation example of the above-mentioned directory creation module is shown in FIG. The directory may be created by a user or an administrator of the system 10 having a hierarchical hierarchical structure in accordance with a specific subject and may be associated with elements such as a search formula, a keyword, a patent classification symbol, , And the system 10 can perform searching and monitoring using the corresponding information, and can perform various analyzes herein on the retrieved or monitored results.

Lower layer patent classification processing module

The lower hierarchical patent classification symbol processing module finds a lower hierarchical patent classification symbol subordinate to the acquired patent classification symbol through the hierarchical patent classification symbol database with respect to the patent classification symbol received from a search formula inputted by the user Role.

In order for the patent document corresponding to the lower-level patent classification code to be included in the search result for the received patent classification code, at least one of the following three is required, and the lower-level patent classification code processing module .

First, the hierarchical patent classification database is used to find all the lower hierarchical patent classification codes of the patent classification symbols obtained based on the time of acquisition. Hierarchical Patent Classifiers Because patent classification symbols on a database have a tree structure, they can be found. When the acquired search formula includes a specific patent classification symbol, all patent classification symbols in the lower hierarchy of the patent classification symbol can be found according to the tree structure, and the result is transmitted to the modified search formula generation module , The modified search expression generation module can generate a new search expression reflecting all transmitted patent classification codes and query the search engine.

Second, if the lower level patent classification codes are found and matched (stored in the matching table) with respect to all patent classification codes in advance, and there is a specific patent classification code in the retrieval formula obtained, Layer patent classification symbols to the modified search formula generation module immediately.

The lower layer patent classification symbol extraction module extracts the lower layer patent classification symbol in two ways. One is a method of extracting only the subclasses of the acquired patent classification code, and the other is a method of extracting all the subclass classification codes that are dependent on the received patent classification code. In this case, even when only the sub-classifier is extracted, the sub-sub-classifier is extracted again for each extracted sub-classifier, and the sub- All sub-patent classification codes can be extracted.

For example, if the patent classification code contained in the search formula is H04B 7/15, the sub-patent classification codes are H04B 7/155, H04B 7/185, and H04B 7/204, The symbols are H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, H04B 7 / 216.

H04B7 / 15, H04B7 / 155, H04B7 / 165, H04B7 / 17, H04B7 / 185, H04B7 / 19, H04B7 / 195, H04B7 / 204, H04B7 / 208, H04B7 / 212 and H04B7 / 216.

It will be appreciated by those skilled in the art that the extraction of such sub-classifications and all sub-classifications may be easily and directly or recursively extracted from the parent-to-parent correspondence or the parent-to-child correspondence It will be natural.

The lower layer patent classification symbol extracting module stores the extracted lower layer patent classification symbol summingly. If the patent classification code included in the above search formula is H04B 7/15, then only H04B 7/15, H04B 7/155, H04B 7/185, and H04B 7/204 ^ H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/15, H04B 7/17, H04B 7/17, H04B 7/15, H04B 7/15, H04B 7/16, H04B 7/16, H04B 7/16, H04B 7/15, 195, H04B 7/204, H04B 7/208, H04B 7/212, and H04B 7/216 ^.

Modified Search Expression Generation Module

The revised retrieval expression generation module generates the revised retrieval expression by integrating the categorized patent classification code with the original existing retrieval expression.

For example, in the case where a given search expression is ^ {keyword = wireless and active} and {applicant = Samsung Electronics} and {IPC = H04B 7/15}, the modified search expression generation module generates IPC H04B 7 / 15 Keyword = wireless and active} and {applicant = Samsung} and {IPC = H04B 7/15 OR H04B 7/155 OR H04B 7 / 185 OR H04B 7/204} or {Keyword = Wireless and Active} and {Applicant = Samsung} and {IPC = H04B 7/15 OR H04B 7/155 OR H04B 7/165 OR H04B 7/17 OR H04B 7 / 185 OR H04B 7/19 OR H04B 7/195 OR H04B 7/204 OR H04B 7/208 OR H04B 7/212 OR H04B 7/216}. Preferably, the latter search expression is preferable.

In summary, when the patent classification code is included in the retrieval formula obtained, the information related to the lower class patent classification symbol of the patent classification symbol is found, and the modified retrieval expression generation module processes I can do it.

If the patent classification symbol included in the obtained search formula includes a dot in the title information, the search engine may search the patent classification symbol preprocessing engine Obtains a sub-patent classification symbol through the patent classification system, binds the obtained patent classification symbol to OR, generates a modified search expression including the sub-patent classification symbol bound to the OR, and performs search using the modified search formula . The above steps are as described above.

Meanwhile, a typical search engine supports range search. For example, in the case of H04B 7/15, H04B 7/15 and H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B It is much faster to query from From H04B 7/15 to H04B 7/216 than to query the search engine by putting together all of the 7/204, H04B 7/208, H04B 7/212, and H04B 7/216 The results can be answered. This is especially true if the number of lower-level patent classification symbols is larger, the difference in the search speed becomes larger.

For the range search, the modified search expression components generated are slightly different depending on whether the search engine or the DBMS supports the disparity range symbol processing for this range. Distinct Range Symbol processing refers to processing a case where symbols on both sides of the range are different, such as ^, which is greater than or equal to ^ A and less than B. In the case of supporting the heterogeneous range symbol processing, it is preferable that, for the obtained patent classification symbol, ^ (patent classification symbols larger than or equal to the obtained patent classification symbol and smaller than the next sibling of the acquired patent classification symbol) Create a modified search expression component ^. (The above-mentioned obtained classification symbol) or (the patent classification symbols that are equal to or larger than the obtained patent classification symbol and smaller than the next sibling of the acquired patent classification symbol) }. Of course, in the case of using a modified patent classification code arranged in a dictionary manner, it is needless to say that the modified patent classification code is processed as described above.

If the retrieval formula obtained is ^ {keyword = wireless and active} and {applicant = Samsung Electronics} and {IPC = H04B 7/15} = Samsung Electronics} and {IPC> = H04B 7/15 and IPC <H04B 7/22}. On the other hand, if the disparity range symbol processing is not possible, ^ {keyword = wireless and active} and {applicant = Samsung Electronics} and {IPC = H04B 7/15 or (IPC> H04B 7/15 and IPC <H04B 7/22) }.

The core of the patent classification symbol preprocessing engine of the present invention is that when there is a given patent classification symbol, all sub-classes (child) of the patent classification symbol or sub-patent classification in the sub-hierarchy are automatically processed, , And surveillance, and the ability to be reflected in the quality consciousness.

When a search result is displayed in a modified search expression, the display of the search result may be provided by clustering according to the searcher's selection. Clustering refers to a method of grouping search results and grouping them into groups, and the present invention is applied as follows. When there is a clustering window representing a clustered structure or when a clustering level can be selected, only the search results of the dot subgroup and subgroups corresponding to the pre-set or pre-selected clustering levels are retrieved, And expresses the search result corresponding to the selected clustering level when the lower or higher clustering level is selected (the clustering level corresponds to the dot level).

For example, when clustering at the 1-dot subgroup level is selected through the search result related to H04B 7/00, the clustering window includes 7/005, H04B 7/01, H04B 7/015, H04B 7/02, H04B 7/14, H04B 7/22, H04B 7/24 appear in the form of a subdirectory folder in H04B 7/00, and the search results are all results that correspond to H04B 7/00 . At this time, when the searcher selects H04B 7/14, H04B 7/145 and H04B 7/15 are displayed in the form of a subdirectory folder. In this case, the search result corresponds to H04B 7/14, Only search results are displayed. If the searcher again selects H04B 7/15, the subordinates H04B 7/155, H04B 7/185, and H04B 7/204 are displayed, and the search results are retrieved according to H04B 7/15 and its subpopulation classification code Only results are displayed.

At this time, for the sake of the searcher's convenience, next to each clustering item (folder name of the directory) in the clustering window, there are several patent classification symbols on its lower right and how many Can be displayed. The number of patent documents corresponding to the patent classification mark and all sub-patent classification marks of the patent classification mark may be displayed. At this time, both the number of documents to be filed and the number of documents to be registered may be displayed, and only the number of documents filed or registered documents may be displayed. For example, if the patent classification code is H04B 7/15, the subtype patent classification code is H04B 7/155, H04B 7/185, and H04B 7/204, so indicate ^ 3 ^ Are H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, H04B 7/216 So, you can display ^ 10 ^. It is preferable to give both quantities, because there are few classifications directly above and below, but there can be huge classifications at the bottom. This is especially true in the emerging new technology sector or in high-tech areas where the rate of technology differentiation or the speed with which detailed classification increases. If only the number of subclassifications is shown, there may be users who ignore the classification because the technical value is less important when the numerical value is small.

Family information preprocessing module

A patent family is a collection of document information consisting of documents directly related to a specific patent document, domestic or overseas. In general, in a single country, the following conditions apply: 1) a split application; 2) an application for change or duplication; and 3) a claim for domestic priority claim (such as some continuation applications and reissue patents in the United States) Which may be slightly different) constitute the domestic patent family. (1) International application (PCT), (2) Treaty priority application, and (3) When the above 1) or 2) has been filed with the Patent Office of a specific country, (1) application for divisional application, (2) application for change or double application, (3) application for domestic priority claim (for some countries such as continuation application and reissue patent in the United States, Etc.) constitute the international patent family. The configuration of the family information preprocessing module for preprocessing the patent family information of the present invention is shown in FIG. In order to process the family information, a family information preprocessing engine 3810 for processing the information is required, and various DBs for obtaining family information are required. Examples of the necessary DBs may include treaty priority information DB 3671, PAJ DB 3673, KPA DB 3675, Inpadoc DB 3677, and other family information DB 3830. In particular, Inpadoc DB (3677) plays an important role.

The family information preprocessing module is closely related to the processing and updating of the patent document mast database. The reason for this is that most of the patent information is 1) a new document is continuously issued, 2) a document is obtained nationwide, and new documents are continuously issued for each country, Or 2) at any time. That is, the family information needs to be updated at the time of obtaining the document of 1) or 2), and the update of the family information at the time of obtaining the document of 1) or 2) is performed by the family information preprocessing module of the present invention. The family information can be found through the existence of a specific application number in the source information or priority information for existing or new documents.

A method of processing the family information by the family information preprocessing engine of the family information preprocessing module is shown in FIG. The family information preprocessing module obtains at least one or more patent document information (S3720), determines whether the patent document information includes family information such as a priority claim number, a split application, a change application, or a continuation application (S3730) If the family information is present, the information of the patent document is stored in the patent database or the search index in association with the unique document number of the family patent document corresponding to the family information (S3740), and the family document (S3750). If there is information about the number of family information in the unique document number of the family patent document, the number is added to the number of the family information (S3750).

The family information processed by the family information preprocessing module is stored in the family information DB of the present invention.

Citation information preprocessing module

In the citation information preprocessing module of the present invention, any one or more of the citation information acquisition module 3400-1 for acquiring information related to the quotation and the cited information update module 3400-2 for updating the citation information based on the citation information . The citation information acquisition module includes an applicant citation information acquisition module 3400-1-1 for acquiring citation related information from the information cited by the applicant in the applicant's patent document and examination citation information There is an acquisition module 3400-1-2. The applicant's citation information acquisition module 3400-1-1 is again connected to the applicant's patent document information acquisition module 3400-1-1-1 and the applicant's non-patent document information acquisition module 3400-1-1- 2). The patent document information acquisition module 3400-1-1 for citing applicant operates when the citation information is patent information. The examination citation information acquisition module 3400-1-2 includes an examiner citation patent document information acquisition module 3400-1-1-1 and an examiner citation non-patent document information acquisition module 3400-1-2. The citation information preprocessing module processes the citation information acquired by the citation information acquisition module in a preset manner.

2) the information of the patents citing the patents as quotation marks cited when the examiner examines specific patent documents; and 3) the indirect reference information citing the patents as indirect references There may be citation information. Of course, it is natural that the above quotation occurs both domestically and internationally. Normally, citing other patents with the above-mentioned prior art is in the prior art literature information in Korea or Japan, and in reference information in the United States. Such prior art citations may be obtained in the bibliography or text of the patent document. In addition, many countries disclose citations for citations, which can be found in administrative information published by the Patent Office. In the present specification, the prior art citation information will be mainly described, but if the obtained historical citation information can be treated as prior art citation information, the above description can be equally applied. The indirect citation includes indirect citation and indirect citation. When there is at least one set of patent documents to be processed, the set of indirectly quoted documents of these patent document sets is a set of indirectly quoted documents of the set of patent documents to be processed, with respect to the patent document to be processed constituting the set of patent documents to be processed, And the patent documents citing the patent document cited by the patent document to be processed simultaneously. The set of indirectly quoted documents of the set of patent documents includes a patent document for citing the patent document to be processed while the filing date of the patent document to be processed is later than the patent document to be processed, Is a combination of patent documents cited at the same time. These indirect citation are often caused by the difference between the patent application date and the patent disclosure period.

As shown in FIG. 5, the citation information preprocessing module includes a preprocessing information preprocessing module for preprocessing the prior art citation information. The preprocessing information preprocessing module preprocesses the prior art citation information and the pre- And preprocessing information preprocessing module for preprocessing the information of the reference citation. Normally, the bibliographic information of the specific document or the backward citation is included in the text, so the document information about the forward reference is stored in the search index or the specific document information of the specific document. However, the forward citation relating to the specific document is not stored in the specific document information but is stored in the search index of the document to which the specific document is cited or in the cited document information. Therefore, it is necessary to search for the backward quotation related to the specific document and store it in association with the specific document, and this function is performed by the preliminary technical citation information preprocessing module.

Prior art citation information preprocessing module

The preliminary technical citation information preprocessing module obtains the rear citation information of the specific document citing the specific document by querying the search index or the DBMS with the unique number value of the specific document or the application number of the specific document . The prior art citation information preprocessing module may include the acquired rear citation information of the specific document in the index in the indexing process of the specific document or may be stored in the patent document master DB as rear citation information of the specific document . While forward citation information is almost unchanged (it is usually stated when the applicant is going to be released, and there may be occasional additions, but the addition of forward citation rarely occurs). The backward citation information is always updated This is the information you need to monitor. Therefore, when a new patent document is generated and the search index is updated or the patent document mast DB is updated, the prior art citation information preprocessing module inquires whether there is an updated patent document citing the specific document, The citation information must be updated.

The citation information preprocessing method for representing the citation number is described in detail in FIG. In step S3820, the preprocessing information preprocessing module obtains at least one quotation number existing in a specific document (S3820), obtains a quotation number that is not a predetermined quotation number attribute in the attribute of the quotation number (S3830) In step S3840, the query number of the predetermined attribute is inquired to the search engine or DBMS in step S3840. In step S3850, the obtained quoted number is represented by the obtained quotation number of the predetermined attribute.

It is necessary to incorporate the backward reference document information into the document information of the specific document after acquiring the backward reference document information which is information on the application document after citing the specific document. Respectively.

The citation information preprocessing module obtains specific document data (S3920), obtains at least one of an application number, a registration number, and a public number included in the specific document data (S3940) (S3950), the back quote information is obtained as the query result (S3960), and the back quote information is included in the bibliographic information of the document data (S3970 ).

Information module for pre-

The pre-citation information preprocessing module performs the same functions as the citation information for the prior art citation information preprocessing module on the prior art citation information.

Counting preprocessing module

The counting preprocessing module will be described. The counting preprocessing module includes a preprocessing module for counting individual document units for performing a counting process in individual document units and a preprocessing module for counting multiple document units for counting processes in two or more document units.

Individual document unit counting preprocessing module

The pre-processing module for counting the individual document unit counts the number of claims, 4) the number of specification pages, 5) the number of documents, 5) the number of patent applicants, The number of patent classification marks, 7) the number of patent classification codes of each kind, 8) the number of references, 9) the number of patent citations, 10) the number of patent classification codes examined by the examiner, The number of patent claims in the country in the reference, 17) the total number of patent documents in the reference, 18) the number of patent claims in the country, At least one numerical information among the number of non-patent documents in the reference can be calculated.

In addition, the above-mentioned individual document unit counting preprocessing module can be used for a period from the filing date to the registration date, (2) the period from the priority date to the filing date, (3) the period from the filing date to the filing date, ) Period from the filing date to the date of the examination request, and 4) period information from the date of notification of the opinion submission to the date of submission of the opinion.

In addition, the individual document count pre-processing module can obtain information on the utilization or application of a specific system for each individual country in the individual patent document or administrative information. The information includes 1) whether or not to use the examination request system, 2) whether or not to use split applications, 3) whether to apply some of the continuation applications, 4) whether to use the reissue patent system, 5) One or more can be identified, and if the system is used or applied, the count value can be set to 1, otherwise, it can be set to zero.

In addition, the individual document unit counting pre-processing module may include information related to the individual patent document in at least one second country (using a search engine or a DBMS) by examining the patent document mast database in association with the individual patent document Can be obtained and counted. Such information may include information about at least one of 1) the number of countries to be filed, 2) the number of countries registered, 3) the number of international applications, 4) the number of families in each country, and 5) .

Also, the pre-processing module for counting individual document units can be classified into 1) forward citation (other domestic and foreign patent documents citing me), 2) backward citation (previous documents that I cite), 3) indirect One or more of the citations can be found and counted.

In addition, the individual document count pre-processing module can obtain information on 1) whether or not to file an objection, 2) whether information is provided, and 3) whether the judgment of a party has occurred. The information on 1) to 3) above can be obtained from the Korean Intellectual Property Office or the Patent Tribunal, either directly or through inquiry methods.

The value counted by the individual document unit counting preprocessing module is preferably stored in the bibliographic database of the patent information mast database. It is further preferable that the counted value is stored in such a manner that the patent document is included in an index generated when the search engine indexes the patent document. The method of including the counting value in the index may include: 1) when the search engine performs indexing, activating the individual document unit counting preprocessing module to obtain a counting value; or 2) counting the count value stored in the patent information mast DB Or a method of acquiring the above information can be used.

The counting preprocessing module refers to the patent classification code mast DB to calculate the number of patent classification codes for each patent classification symbol by 1) the total number of sub-patent classification codes, 2) the number of sub-patent classification codes, It can be stored in the mast database.

The manner in which the counting preprocessing module handles counting is well illustrated in FIG. First, the counting preprocessing module obtains at least one patent document information (S2820), calls at least one or more individual document unit pre-processing modules for each patent document acquired (S2830), and the individual document unit count pre- (S2840). The counting result is stored in the patent DB or the search index together with the document unique number of the patent document information (S2850)

Multi-document unit counting preprocessing module

Hereinafter, a module for acquiring the score calculated in units of the plurality of documents, i.e., a multi-document unit counting preprocessing module will be described. The preprocessing module for counting multiple document units obtains a score by collecting information counted by the individual document unit pre-processing module in a plurality of documents constituting the target set. That is, the multi-document unit counting preprocessing module generates counting data for at least two document sets.

Weight pre-processing module

Document unit weight preprocessing module

It is impossible to systematically measure the value of inventive ideas contained in a patent document. However, it would be irrational to treat all patent documents equally, as it is impossible to measure the value accurately. Therefore, the present invention discloses a method for weighting patent documents based on information that the system can grasp as data. The document unit weight pre-processing module of the present invention is responsible for the weighting function of the patent document.

The weight pre-processing module of the present invention includes a document unit weight pre-processing module 3310 and a subject unit weight pre-processing module 3330. The document unit weight preprocessing module 3310 includes a cost expenditure perspective weight preprocessing module 3311, a citation perspective weight preprocessing module 3313, a conflict perspective weight preprocessing module 3315, and / or a central perspective weight preprocessing module 3317 . &Lt; / RTI &gt; The subject-based weight pre-processing module 3330 may include an applicant unit weight pre-processing module 3331, an inventor-based weight pre-processing module 3333, and / or a proxy unit weight pre-processing module 3335.

Expenditure perspective weighted pre-processing module

From a probabilistic, statistical or societal point of view, it is likely that more costly patents are more important. This assumption is highly likely to be established from the perspective of rational allocation of costs in the same entity, and this assumption is likely to be established among other competitors. The following weighting factors can be considered in the structure where a patent is paid for a patent.

(1) The number of claims, the number of independent claims, the number of pages (number of pages), the number of families, or the number of patent classification symbols in quantitative terms. 2) Patents pending, registered, rejected 3) the number of applicants or inventors from a subjective point of view; and 5) the number of applicants from a subjective point of view. 5) Factors such as the use of international applications from an international perspective and the number of countries entering the foreign country stage may affect the value of patent documents. These weighting factors can be obtained from bibliographic information of individual documents, specification content information or administrative processing information.

In general, if the number of claims is large or the amount of specification is large, the contents of the invention may be wide and diverse, and it is highly likely that the inventions are time-consuming and costly, and agents are more likely to charge more. In addition, the greater the number of families or the greater the number of patent classification symbols, the more likely it is that the invention is comprehensive. In addition, it is more likely that the registered patent is more important than the pending patent, and it is more likely that the patent application filed for the examination request or the application filed for the priority examination is more important than the patent application filed for the non-patent application. If the applicant is more than 2, it is a joint application. Therefore, it is more likely to be statistically significant than the other cases because it is the result of cooperation between two or more subjects. Even if there are several inventors, It is likely to be statistically significant. In addition, since the overseas application is usually much more expensive than the domestic application, the applicant has a considerable financial expenditure when there is an overseas application or an overseas application, so that the patent is relatively important to the applicant It is reasonable to judge it to be high.

From the above viewpoint, it is possible to grasp the existence and numerical value of a weighting factor for one document (if there is no problem only if there is no problem, the numerical value will be 0, if there is 1, And a weight score may be assigned to each of the weighting factors according to the existence and / or the numerical value. In order to grasp the existence and numerical values of the detailed weighting factors constituting each aspect of the above 1) to 5) with respect to one document and to assign a weighting score to each weighting factor, the cost expenditure perspective weighting pre- .

The manner in which the cost-expenditure-based weighting pre-processing module processes the weights is well illustrated in FIG. The cost expenditure perspective weight pre-processing module obtains at least one or more patent document information (S2920), and uses the counting preprocessing module or the patent document information obtained through the patent DB or the search index, A counting result value for each counting criterion is obtained (S2930), and a weight for each counting criterion is obtained from the patent information processing policy DB (S2940). At this time, information on the predetermined counting criteria is referred to the patent information processing policy stored in the patent information processing policy DB. For example, in view of the claims of the patent information processing policy DB from the viewpoint of cost expenditure, the policy for processing weights is stored for each subject in each viewpoint (claim count * 0.1 + independent count * 0.3). The cost-expenditure-based perspective weighting preprocessing module generates a cost-based weighted value as a result of counting for each counting criterion and each counting criterion (S2940), compares the generated weight value with the unique document number of the patent document information, In the search index (S2950).

Citation perspective weighted pre-processing module

As with the paper, it is highly probable that many patents cited by the majority are important patents. A module that processes the weights of individual documents from this citation perspective is referred to as a citation perspective weighted preprocessing module. The point of view of the quotation is divided in detail: 1) the number of total rear citation, the number of rear citation, the number of backward citation in the first stage, the average period of rear citation, 2) There may be weighting factors such as whether the examiner is cited in terms of the examiner's citation. In other words, the more the citation of the patent is directly or indirectly quoted (the patent cited by the patent is quoted by the other patent), the more the first step citation is directly cited, The more the authors cite their own patents, the more likely it is that they will become more important. In addition, patents quoted by the examiner during the review process may be of greater importance than those that are not. The way in which the citation-based view weight pre-processing module preprocesses the weights in terms of citation is similar to that of the above-mentioned cost expenditure perspective weight pre-processing module, as shown in FIG.

The citation-based view weight pre-processing module obtains at least one or more patent document information (S3020), and uses the counting preprocessing module, the forward citation and / or the rear citation or the examination citation in the patent document information obtained through the patent DB or the search index (S3030), obtains a weight value for each counting criterion by referring to the patent information processing policy DB (S3040), and calculates a weight for each counting criterion and a counting criterion (Step S3050). The generated weight value is stored in the patent DB or the search index together with the unique document number of the patent document information (S3060). At this time, Information about the counting criteria is referred to the patent information processing policy stored in the patent information processing policy DB. For example, the patent information processing policy DB from the above cited view stores a policy for processing weights for each subject in each view such as "count of rear citation count * 0.3 ".

Dispute perspective weighted pre-processing module

The more disputes a patent has, the greater the importance of the patent. The disputed elements that can be identified in patent information are: 1) invalidation, 2) objection, 3) information provision, and 4) passive or active rights judgment. The dispute perspective weighting pre-processing module of the present invention performs the determination of the presence and the numerical value of the weighting factor from the disputed viewpoint and assigns a weighting score to each weighting factor.

The manner in which the dispute perspective weighting pre-processing module processes the weights is illustrated in FIG. 81, which is similar to the manner in which the cost expenditure perspective weight pre-processing module and the citation perspective weight pre-processing module process weights.

The dispute perspective weight pre-processing module obtains at least one patent document information (S3120), and uses the counting preprocessing module or the patent document information obtained through the patent DB or the search index to generate at least one predetermined counting criterion (S3130). Then, the weighting value for each counting criterion is obtained by referring to the patent information processing policy DB (S3140). The weighting value for each counting criterion and the counting result value for each counting criterion (S3150), and stores the generated weight value in the patent DB or the search index together with the unique document number of the patent document information (S3160).

That is, the dispute perspective weight pre-processing module processes the weights by counting the number of disputes and obtaining the weight for each count number from the patent information processing policy database to generate a dispute point weight value do.

Concentration perspective weighting preprocessing module

It is highly probable that the technology field in which various subjects are concentrated based on a specific point in time is an important technology field. One applicant rarely evenly applies to various technical fields. He or she applies to a place where he or she thinks it is important, places where technology investment is concentrated, and places a defensive minority in defenseless places Most of them do. Therefore, the technical field to which one application belongs is that many applicants have a lot of applications compared to other technical fields, 1) the number of applications is high, the rate of increase or increase of application is high, 3) If the change in market share is out of the predetermined range, then the technology field is likely to be an important technology field. The areas in which the various applicants concentrate can be measured by such analytical factors as technology attractiveness, rate of increase, rate of increase, change in market share. At this time, it is desirable that the technical field in this paragraph grasp the layered patent classification symbols included in patent documents such as IPC, USPC, FI, FT, ECLA, etc. in each layer unit. Of course, it would be more desirable to automatically include the subpopulation classifier to calculate the numerical value for each analysis element. In other words, when IPC is H04B 7/26, it is possible to calculate the numerical values of various analysis elements for this IPC, and also to calculate the numerical values of various analysis elements from the viewpoint of H04B 7/00. At this time, it is more preferable that other patent classification symbols belonging to the sub-patent classification symbol H04B 7/26 are hierarchically included. From the viewpoint of concentration, the concentration-weighted weight pre-processing module of the present invention performs the identification of the presence and the numerical value of the weighting factor and assigns a weight score to each weighting factor.

The manner in which the convergence-based weighting pre-processing module processes the weights is contrary to the manner in which the cost-expenditure-based weighting preprocessing module and the citation-based weighting preprocessing module process the weights. That is, the bibliographic information for each individual document measures the value of the patent information analysis index including the concentration of the applicant / patentee in terms of concentration / activity / market share in the technical field, and assigns a weight to each value And obtains the concentrated view weight value from the information processing policy DB.

The manner in which the centralized point weighting pre-processing module processes the weights is illustrated in FIG. 82, which is similar to the manner in which the cost-expenditure-based weighting preprocessing module and the citation-based weighting preprocessing module process the weights.

The centralized point weighting preprocessing module obtains at least one patent document information (S3220), obtains the applicant information and the patent technical classification information from the patent document DB (S3230), and obtains the applicant information and the patent technical classification information as values Referring to the patent technical classification code DB to calculate the value based on at least one of the patent technical classification obtained on the patent technical classification code system or the applicant's concentration / activity / occupancy rate of at least one higher patent technical classification code (S3240), weight information on the concentration / activity / occupancy ratio of each criterion is obtained from the patent information processing policy DB (S3250), and a concentrated perspective weight value is generated from the criterion weight value and the concentration result value (S3260) , The generated weight value is stored in the patent DB or the search index together with the unique document number of the patent document information (S3270).

The concentrated-point weighting preprocessing module may process weights in units of inventors rather than applicants, and the manner in which the weights are processed is shown in FIG. The centralized point weighting preprocessing module obtains at least one or more patent document information (3320), obtains the inventor information and the patent technology classification information from the patent document information (S3330), obtains the inventor information and the patent technology classification information , Referring to the patent technical classification code DB to calculate the concentration of the inventor on the patent technology classification or at least one or more higher patent technical classification codes obtained on the patent technical classification code system in step S3340, (Step S3350). The centralized weight value is generated from the weight value of the reference and the concentration result value (S3360), and the generated weight value is added to the patent database or the search index (S3370).

Subject-unit weight pre-processing module

When there is a patent document, if the subject involved in the patent document is an important subject, the patent applied by the subject is likely to be an important patent. The key is to judge which subject is more important. A subject that includes a patent document may be an applicant, an inventor, and an agent. A preprocessing module for subjecting a patent document to the function of preprocessing the weight of the document from the subject point of view. As shown in FIG. 4, the subject-based weight pre-processing module includes an applicant unit weight pre-processing module, an inventor unit weight pre-processing module, and an agent unit weight pre-processing module, depending on the subject.

Applicant unit weight pre-processing module

If the technical field included in the specific document is the technical field to which the applicant of the document is concentrated, the importance of the application is likely to be significant. The field in which the applicant of a specific document focuses can be measured by analysis factors such as concentration rate, market share, and AI in the technology field. The unit weight pre-processing module, which is an applicant of the present invention, performs numerical values of the weighting elements from the viewpoint of the above-described analysis elements and assigns weighting scores to the respective weighting elements.

The manner in which the Applicant's unit weight pre-processing module processes the weights is shown in FIG. The applicant's unit weight pre-processing module obtains the reference information for processing the weights in units of applicants (S3420), and obtains the result values calculated by at least one document unit weight preprocessing module for the documents of applicant's name conforming to the standard (S3430), weight information for each document unit weight preprocessing module is obtained from the patent information processing policy DB (S3440), and a weight value of each document unit weight preprocessing module and an applicant unit weight value obtained by each result value are generated (S3450). The created weight value is stored in the patent DB or the search index together with the reference or independently of the patent document information (S3460).

Inventor unit weight pre-processing module

If the technical field included in a specific document is a technical field concentrated by the inventor of the document, the importance of the application will be high. The field in which the inventor of a particular document focuses may be measured by an analysis factor such as concentration ratio, occupancy, and AI in the inventor's standard in the technology field (the formula is the same as when the inventor was substituted in the applicant reference formula). The inventor's unit weight preprocessing module of the present invention grasps the numerical value of the weighting factor from the viewpoint of the above-described analysis factor and assigns a weighting score to each weighting factor.

The manner in which the inventor's unit weight pre-processing module processes the weights is shown in FIG. The inventor's unit weight pre-processing module obtains the reference information for processing the weights in units of inventors (S3520), obtains the result values calculated by at least one document unit weight preprocessing module for documents of the inventor's name conforming to the standard (S3530), weight information for each document unit weight preprocessing module is obtained from the patent information processing policy DB (S3540), and a weight value of each document unit weight preprocessing module and a criterion value (S3550). The created weight value is stored in the patent database or the search index together with the reference or the unique document number of the patent document information (S3560). The weight information may include a policy such as "registration rate of past total application * 0.1 + number of overseas filing families * 0.5".

Agent unit weight pre-processing module

If the technical field included in a specific document is a technical field that the agent of the document focuses on, the completeness of the patent specification will be higher than that of a non-representative. Therefore, the field in which the agent of a particular document focuses can be measured by analysis factors such as concentration, share, and AI (the formula is the same as when it was replaced by the agent in the applicant reference formula) will be. The agent-unit weight pre-processing module of the present invention performs the numerical value of the weighting element from the viewpoint of the above-described analysis element and assigns a weighting score to each weighting element. The weight of the agent unit may be lower than the weight of the applicant unit or the weight of the inventor unit.

The manner in which the proxy unit weight pre-processing module processes the weights is shown in FIG. The proxy unit weight pre-processing module obtains the reference information for processing the weight in the proxy unit (S3620), and obtains the result value calculated by at least one document unit weight preprocessing module for the document of the proxy name matching the standard (S3630), weight information for each document unit weight preprocessing module is obtained from the patent information processing policy DB (S3640), and weight values of each document unit weight preprocessing module and a representative agent unit weight value obtained as each result value are generated (S3650). The created weight value is stored in the patent DB or the search index together with the reference or the unique document number of the patent document information (S3660). The weight information may include a policy such as "registration rate * 0.3 relative to application ".

The document unit weight pre-processing module and the subject unit weight pre-processing module according to the present invention assign a predetermined weight to the weighting factors acquired by at least one or more weight pre-processing modules as described above, and calculate the weights of document units. At this time, each of the weighting modules of each viewpoint or each of the weighting modules for each subject can give a predetermined weight to the weighting factor acquired by itself, and calculate the weight of the document unit from the viewpoint of the user. It will be appreciated that each of the above weights may be assigned differently depending on the input of an administrator or a user.

The subject mast DB generation module 301-4

Type of Subject

As for the patent document, the subject may be 1) an applicant, 2) an inventor, and 3) an agent. The applicant may be an organization and an individual such as an enterprise. In the present invention, a DB for storing and managing the subject is referred to as a subject mast DB 204, and a module for generating the subject mast DB 204 as described above is referred to as a subject mast DB generation module 301-4.

The subject matter mast DB 204 may further include a representative name DB, a well-known inventor name DB, and a representative name of a representative agent DB, and may further include an enterprise information DB independently. These will be described separately.

The configuration of the subject mast DB generation module 301-4

The subject mast DB generating module 301-4 of the present invention includes the representative name pre-processing module 301-4-1. The representative linguistic pre-processing module includes a representative linguistic pre-processing module (301-4-1-1), an inventor representative linguistic preprocessing module (301-4-1-2) for processing the inventor's representative linguistic name, an agent And representative representative linguistic pre-processing module 301-4-1-3 for performing representative linguistic processing. Meanwhile, the subject mast DB generating module 301-4 of the present invention may further include a company information DB generating module. The corporate information DB may further include at least one of financial information of company, information of company status, accounting information, main product, representative, corporation registration number, business number, website, telephone number and fax number. The subject matter mast DB 204 is further provided with an enterprise information DB 204-1, a representative name DB 204-2, a representative name DB 203-3, a representative name DB 204-4, and / or an organization information DB 204-5.

Representative phrase extraction pre-processing module

Concept of representative phrases

Next, the representative word extraction module of the present invention will be described. First, the concept of representative phrases is explained. A Representing Phrase is a word consisting of two or more words that can represent a whole or a part of a specific document through the representative phrase. The basic attributes of the representative phrases of the present invention are a set of predetermined representative phrases extracted from a given set of representative phrases and a set of representative phrases (including one representative phrase extraction objective document) A part of the objective document or a common representative part of each document of the predetermined representative phrase extraction target document set frequently comes out above the predetermined standard.

The set of predefined representative phrase extraction whole document sets includes, for example, 1) a set of patent documents generated by a method of specifying a set of patent documents with a specific patent classification symbol on a specific patent classification system in a specific country DB, 2) (3) the name of the specific inventor (ie, the name of the applicant and the inventor) included in the patent document of the specific applicant in the specific country DB as the inventor 4) a set of patent documents generated by a method of specifying a set of patent documents by a specific agent name in a specific country DB, 5) a set of patent documents generated by a method of specifying a set of patent documents by a specific country DB A specific patent classification symbol on a specific patent classification system and a patent document set by a specific applicant 6) a patent document generated by a method of specifying a set of patent documents by the name of a specific applicant and a specific inventor by a specific patent classification symbol in a specific patent classification system in a specific country DB, 8) a set of patent documents generated by a method of specifying a set of patent documents by the name of a specific applicant and a specific agent in the specific country DB, 9) a set of all patent documents of a specific country, 9) A set of all patent documents, and 10) a set of patent documents specified in units of a predetermined period in the above 1) to 9).

The representative phrase extraction target document set includes at least one or more specific attributes (for example, a specific patent classification code such as IPC, applicant, period, country, etc.) from an arbitrary subset of the representative phrase extraction whole document set An agent or inventor, or a combination of one or more of these) may be a subset of documents in common.

All.

The representative set extraction target document set (the super set including the target set) is a concept relative to the representative document extraction whole document set (target set), and the subset of the representative document extraction whole document set is sufficient. Preferably, The smaller the size of the set, the better. Examples of the representative phrase extraction target document set include 1) a specific application, all applications of a specific company, an application by a specific inventor, a specific IPC application by a specific inventor, all applications of a specific IPC of a specific company, All of which are examples.

Examples of a part of a representative phrase extraction objective document include an example of a patent claim, an independent item in the claims, a subordinate item in the claims, an effect of the invention, industrial applicability, and a summary. On the other hand, a common part of each document of the predetermined representative phrase extraction target document set will be the above-mentioned patent claim range of all the documents constituting the representative phrase extraction target document set.

The representative phrases include 1) a phrase consisting of only nouns, 2) a phrase consisting of only nouns and verbs, and 3) a phrase consisting of adjectives and adverbs. The length of the representative phrase is preferably 1 to 5, more preferably 2 to 3, based on the number of words. Typically, a technical concept is created in one word, but in many cases it is often the case that a technical concept or a technical action or effect consists of two or three words.

The representative phrase extraction target document set is a relative concept to the representative phrase extraction whole document set. The subset of the representative phrase extraction whole document set is sufficient. Preferably, the smaller the subset size is, the better. Examples of the representative phrase extraction target document set include 1) a specific application, all applications of a specific company, an application by a specific inventor, a specific IPC application by a specific inventor, all applications of a specific IPC of a specific company, All of which are examples.

Examples of a part of a representative phrase extraction objective document include an example of a patent claim, an independent item in the claims, a subordinate item in the claims, an effect of the invention, industrial applicability, and a summary. On the other hand, a common part of each document of the predetermined representative phrase extraction target document set will be the above-mentioned patent claim range of all the documents constituting the representative phrase extraction target document set.

Next, the representative phrase extraction process will be described in detail. The representative phrase extraction is performed by the representative word extraction preprocessing module of the present invention. The representative word extraction preprocessing module of the present invention includes various dictionary databases including 1) a morphological analysis engine for each language, 2) a phrase generation engine, 3) a classification counting engine, and 4) a thesaurus dictionary / , 5) a representative phrase extraction engine. The configuration of the representative phrase extraction preprocessing module is well illustrated in FIG. A representative phrase extraction preprocessing engine 3710 including an algorithm for extracting representative phrases, a representative phrase extraction policy DB 3730 including information on a policy as a basis for extracting representative phrases, A representative word-frequency-application number correspondence DB 3750 including information on various counting or calculation information such as the frequency value for each representative word, and information on which position (field, etc.) of the document the representative word comes from . On the other hand, a thesaurus DB 3770 for processing synonyms and synonyms, and a translation system for processing two or more languages, or a representative phrase translation DB including translation information for representative phrases using a previously translated dictionary (3790) may be further included. Hereinafter, the representative phrase extraction preprocessing engine of the present invention will be described in more detail. The representative phrase extraction preprocessing engine performs a function of extracting a terminology when a sentence or a paragraph of a specific field is input.

The phrase generation engine of the present invention generates 1 to 5 (preferably 2 to 3) phrases in units of phrases. For example, when a sentence included in a patent document is input, let the output phrases be a b c d e f, g in order. At this time, if the phrase generation unit of the phrase generation engine is three words, the phrase generation engine generates abc, abd, bcd, cde, def, and efg. The phrase generation engine generates a phrase on a sentence unit basis with respect to a given document or document portion. An example of creating a phrase for a document part may be to generate the phrase for the sentences in the claims. The method of generating the arrangement is as follows: 1) a phrase is generated by using nC2, nC3, nC4, and nC5 for the n phrases when n words are extracted from one sentence, For example, when nC2 is applied, the digits ab, ac, ad, ae, af, ag, bc, ... fg are generated.) 2) B, cd, de, ef, fg, and so on) when the two arrangements are sequentially arranged. , And abc, bcd, cde, def, and efg when three phrases are arranged.) Of course, you can use both arrays of two and three phrases.

On the other hand, the phrase that has passed through the morpheme analyzer may also be called a term. The term may vary depending on the setting value in the morpheme analyzer. In this case, examples of the setting values include 1) separating only the vowel, 2) extracting only the representative word, but extracting only the representative word, and 3) extracting only the representative word in the case of the adjective or adverb.

Then, the erroneous counting engine essentially generates information (a phrase, a document number) or (a phrase, a counting value) for each of the generated phrases, and preferably generates information (a phrase, a counting value, a document number) (E.g., a phrase, a counting value, a document number, a location field name on the document (for example, a patent claim), a location field name on the document number ) Information. The distinguishing counting engine stores the generated information in the representative word DB 207-2. Table 5 shows an example of data included in the representative word DB 207-2. The example shown in Table 5 below shows an example of data generated for specific phrases by country.

[Table 5]

Fish
ID Fish Word count Cumulative count value in document (based on document number) country code Article number Fishing location ... "False absolute ID" One abc 3 One KR 10-2003-0012345 Patent Claims One One abc 3 2 KR 10-2003-0012345 Patent Claims 2 One abc 3 3 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 3 One abc 3 4 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 4 One abc 3 5 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 5 One abc 3 6 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 6 One abc 3 7 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 7 2 bcd 3 One KR 10-2003-0012345 Patent Claims 8 2 bcd 3 2 KR 10-2003-0012345 Patent Claims 9 2 bcd 3 3 KR 10-2003-0012345 Patent Claims 10 2 bcd 3 4 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 11 2 bcd 3 5 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 12 2 bcd 3 6 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 13 2 bcd 3 7 KR 10-2003-0012345 DETAILED DESCRIPTION OF THE INVENTION 14 ... ... ... ... ... ... ... ... One abc 3 One KR 10-2003-0056789 Patent Claims 150 One abc 3 2 KR 10-2003-0056789 DETAILED DESCRIPTION OF THE INVENTION 151 One abc 3 3 KR 10-2003-0056789 DETAILED DESCRIPTION OF THE INVENTION 152 ... ... ... ... ... ... ... ...

It can be seen that the data shown in Table 5 is generated by the patent documents 10-2003-0012345 and 10-2003-0056789, and it can be seen that the phrase generation engine has generated the phrases in three units. In the above-mentioned 10-2003-0012345, it can be seen that the phrase abc has been issued twice in the claims and five times in the detailed description of the invention. Also, it can be seen that the phrase bcd is three times in the claims and four times in the detailed description of the invention. In addition, the phrase abc can be found in the patent document 10-2003-0056789 once in the claims and twice in the detailed description of the invention. Therefore, the categorical counting engine of the present invention can generate data as shown in Table 5 for all patent documents repeatedly or recursively obtained. In addition, the phrase generation engine may generate a phrase in two units, and the word counting engine of the present invention can generate the data as shown in Table 11 for phrases generated in the two units It will be self-evident. Further, even in one or four or five units, the above-mentioned counting engine will be able to generate the data shown in Table 5 above.

At this time, the generation of the distinction ID may be a problem. The method of generating the distinction ID is obvious in the conventional DB technology, but presents some exemplary methods. In the first method, the IDs are sequentially given to the first generated phrases, the existing phrases are searched for the next generated phrases, the new IDs are assigned if there is no phrases, and the existing IDs are assigned if they exist. This process is repeated or recursively performed for every available phrase. In the second method, temporary IDs are sequentially assigned to all the phrases obtained without the IDs for the phrases to generate the data as shown in Table 11, and the same formal IDs are assigned to the same phrases while reading one for each phrase, This process is repeatedly or recursively performed on phrases for which no formal ID is assigned.

The discrimination counting engine of the present invention can generate a phrase absolute ID every time a phrase is generated and processed and assign it to every phrase.

In addition, the inventive counting engine of the present invention can generate each cumulative counting value based on the target document set. The target document set may be counted based on one document as shown in the above table, and it is obvious that cumulative counting can be performed even in a specific field (for example, patent claim) in one document.

The erroneous counting engine of the present invention can generate a cumulative counting value based on the entire document set and store it in the manner shown in Table 6. In the above example, 10-2003-0012345, and 10-2003-0056789, the numerical values are entered based on the above table, and the present invention is actually applied to a much larger overall document set.)

[Table 6]

Fish
ID Fish Cumulative counting value (based on total document set) Absolute ID of related phrase ... ... ... ... ... ... ... ... One abc 10 1, 2, 3, 4, 5, 6, 7, 150, 151, 152 2 bcd 7 8, 9, 10, 11, 12, 13, 14 ... ... ... ... ... ... ... ...

In addition, the segregated counting engine of the present invention can generate a counted cumulative counting value as shown in Table 7 on the basis of a specific field on the document (for example, claims).

[Table 7]

Fish
ID Fish Cumulative counting value (based on claims) Absolute ID of related phrase ... ... ... ... ... ... ... ... One abc 3 1, 2, 150 2 bcd 3 8, 9, 10 ... ... ... ... ... ... ...

......... The representative phrase extraction engine of the present invention will be described. Representative Phrase Extraction If there is data (phrase, document number) data extracted from the entire document set, it will be easy to generate data (phrase, counting value, document number) data as shown in Table 8 below. The above (phrase, counting value, document number) data may be generated by increasing the counting value by 1 when there is the same phrase. The representative word extraction engine of the present invention can generate (a phrase, a counting value, a document number) based on data (a phrase, a document number) and a position field name on a phrase, a document number, ) Data, (phrase, counting value, document number, location field name on document number) data. The above process can be performed not only for all documents belonging to at least one representative phrase extraction target document set but also for at least one representative phrase extraction whole document set that is the sum of the representative phrase extraction target document set You can do it.

The representative phrase extraction engine is an example of a representative phrase extraction set of all documents: 1) all documents belonging to one national DB, 2) all documents corresponding to at least one given patent classification, 3) all documents for a specific applicant, 4 ) All documents for a specific inventor, or (5) all documents produced by limiting the duration of each of these, or (6) all of the documents generated by combining them (including any aggregate operations such as union, difference or intersection) (Eg, phrase, document number) data or a (location field name on the phrase, document number, and document number) data for the document. On the other hand, data (a phrase, a document number) or data (a phrase, a document number, and a location field name on the document number) data may be generated for all documents corresponding to an arbitrary subset of the representative document set.

At this time, the representative phrase extraction engine extracts at least one representative phrase from the at least one or more representative phrase extraction sets for each set of at least one representative phrase extraction target document set for each of the phrases 1) frequency in the representative phrase extraction target document set (frequency F ) And 2) the total frequency (T) in the representative document set extracted from the representative phrase extraction set which is the sum of the collection of the representative document extraction target document set. In addition, it is possible to calculate the total frequency (A) of the representative text in the representative text extraction set, and calculate the total frequency of the AT in the representative text extraction target document set Of course I will.

At this time, the representative phrase extraction engine refers to the representative phrase extraction policy DB and extracts a representative phrase of the representative target phrase extraction target document set from the given representative phrase extraction whole document set. At this time, the representative phrase extraction policy may be whether or not the probability value of the phrase is included in a preset reference range under a predetermined condition (probability value of the phrase, phrase). Examples of the probabilities may be F / T, F / A, F / AT values for specific phrases. The representative phrase extraction policy is defined as follows: 1) Variation value (increase / decrease rate) of F, T, A, and AT values; 2) Variation (increase / decrease rate, increase / decrease rate) of F / T, F / It can be instructed to satisfy the predetermined reference range. At this time, the predetermined condition of the phrase extraction policy may be determined in accordance with the above (the probability value of the phrase and the phrase) by the period, the country, the applicant, the inventor, or at least one patent classification symbol, The probability value of the phrase may correspond to a value of another reference range. For example, the probabilities for the phrases extracted in the IPC H section with many patent applications and the probabilities of the phrases extracted in the IPC D section may be applied differently. Of course, the above-mentioned phrases are also extracted in the values of F / T, F / A, F / AT fluctuation (increase / decrease rate, increase / decrease rate) Different criteria may be applied flexibly according to the attributes of the document set (representative phrase extraction whole document set, representative phrase extraction target document set).

Hereinafter, the concept of representative word extraction of the present invention will be described with a brief table. The representative phrase extraction engine of the present invention is configured such that 1) a probability value (= number of occurrences of a specific phrase / total number of phrases) is associated with each phrase ID on the basis of frequency of occurrence of all phrases, and phrases corresponding to a predetermined probability value range And the representative phrase for each target document set is selected by referring to the predetermined representative phrase selection rule. It will be appreciated that the number of occurrences can be calculated for each specific field. In the example shown in Table 8 below, the number of occurrences of each word can be counted according to the document number, and a representative word can be extracted for each document number conforming to the predetermined representative phrase extraction policy.

[Table 8]

Old I fish The number of occurrences in the representative document extraction target document set (F) (frequency) Article number Representative phrase extraction Number of occurrences in total document set (T) F / T (%) Number of occurrences / Total number of occurrences (%) Number of occurrences in total T / A / Total number of total words (1 billion) (%) Total number of fishes (A) abc 40 #One 80000 0.05 0.008 1,000,000,000 abc 2 #2 80000 0.0025 0.008 abc 200 # 3 80000 0.25 0.008 bcd 8 #One 200 4 0.00002 bcd 100 #2 200 50 0.00002 bcd 2 #4 200 0.1 0.00002 cde 15 #One 3000 0.5 0.0003 cde 100 #4 3000 3.3 0.0003

As shown in Table 8, there will be a phrase absolute ID corresponding to 1: 1 for every phrase. Therefore, the data (phrase, document number) is essentially equivalent to the data (absolute ID, phrase, document number). A phrase ID corresponds to each phrase absolute ID, a document number from which the phrase is associated for each phrase absolute ID, and a bibliography is associated with each document number. Therefore, obtaining the number of occurrences of the document ID and the document number of each of the document IDs in Table 8 above may be an example of the target document set as the document number. Similarly, a predetermined level IPC may correspond to the document number place, a specific applicant or a specific inventor may be corresponded, and in each case, the number of occurrences of each phrase ID per each target document set may be counted. The DBMMS can count the number of occurrences by the ID of each target document set through the table-specific join command.

In this case, for example, if the one-way rule constituting the representative phrase extraction policy in the representative phrase extraction policy DB is 1) the number of occurrences (T) in the total number of phrases (for example, 1 billion) 2) the number of occurrences in the target document number is 1 / 1,000,000% or more; and 3) the number of occurrences / total number of occurrences (%) is 10% to 1% Of the total number of words. The phrase abc can not be a representative word in any of document numbers # 1 to # 4 (violation of condition 1), and the phrase bcd can be a phrase representing the document number only in document number # 2, And abbreviation for document number 4), the word cde can only be a representative word for document number # 4 (document number 1, condition 3). Therefore, bcd and cde are not included, and the representative word of the document number # 2 includes bcd. The document abc is not included in the document number # 3, and the word cde can be included in the document number # 4.

In this way, the representative word extraction preprocessing module of the present invention can generate categorical counting data by using the indexer 401-3 of the search engine and the index, and based on the categorical counting data, The representative phrases can be extracted from the contents of each specific field extracted from documents, predetermined document sets, specific field contents, or a predetermined document set in the same manner as in FIG. For example, in the upper paragraph, the document number # 2 can correspond to the bcd in the representative phrase. If representative phrases are extracted for each set of representative phrase extraction target documents, the representative phrases may be stored in units of the representative phrase extraction target document set or individual document units constituting the representative phrase extraction target document set. That is, at least one representative phrase may correspond to one representative phrase extraction target document set unit or the individual document. The representative word may be corresponded to at least one of individual field units of the representative word extraction target document set or individual field units of the individual documents constituting the representative word extraction target document set. For example, the phrase bcd can be stored in correspondence with the document number # 2. At this time, when the # 2 is called, the phrase bcd can be displayed.

At this time, too many representative phrases can be assigned to specific document numbers. That is, for a specific document, there may be 100 representative phrases satisfying the above-mentioned conditions, and in this case, 4) 10 to 30 representative phrases in the order of frequency in which each representative phrase is calculated, , The number of representative phrases can be limited. On the other hand, there may be a case where there is not one representative phrase for a specific document. In this case, condition 3) may be relaxed in the above-mentioned condition so that the number of representative phrases can be maintained at 10 to 30 .

If the above process is repeatedly performed on all documents or all sets of documents, a representative phrase of a predetermined number range (for example, 10 to 30) may be associated with all documents. Therefore, the representative word corresponding to the document number can be obtained through the process described above. The calculation values (for example, the number of occurrences of the document number in the document, the number of occurrences in the whole document, and the like) for the representative word and the representative word are treated as an example of the bibliography for the document number, Can be utilized.

In the above embodiment, the appearance frequency is calculated in units of documents. The appearance frequency is the following: 1) a document set or 2) a field unit within the document (claim scope etc.) or 3) (Eg, paragraph 1 of the patent claims included in all documents in IPC H04B 7/02 of the Republic of Korea filed between 2000 and 2005). At this time, in the case of one document, it can be handled in the same way as one document (the corresponding point of the document number). However, in the case of two or more sets of documents, the correspondence of document numbers may seem to disappear, but one of the following three methods can easily solve this problem. (1) a method of extracting representative phrases as a unit of individual documents belonging to a document set, then performing a union operation of the representative phrases, and then limiting the number of representative phrases to a predetermined range (for example, 10 to 30) , 2) if representative phrases have already been extracted for each document number, obtaining a representative phrase corresponding to the document number belonging to the document set, performing a union operation, and then limiting the number of representative phrases to a predetermined range; ) After processing all the documents belonging to the document set, one of the methods of extracting the representative words corresponding to the unique ID of the document set other than the document number may be used for each of the words. If the content unit is included in the field of all the documents in the above 3) document set, the above 1) summation processing and 2) document unique ID correspondence processing method may be used. If a representative phrase is provided in the field unit, The above-mentioned 3) method of collecting after acquisition may be used.

Hereinafter, a method of extracting the representative word using an indexer 401-3 (indexer) of the search engine connected to the morpheme analysis engine will be described. The reason why the search engine is used is that the search engine typically has excellent performance in counting the number of search word values, and can easily convert data indexed by the indexer 401-3 into data for DB Because. The indexer 401-3 stores the obtained term and the document number from which the term comes, or, in some cases, the field name from which the term appears. For example, (term 1, # 1) means that term 1 is obtained from document # 1, and (CL: term 1, # 1) means that term 1 is a claim ) In the field. The index of the search engine contains a large amount of data such as (term 1, # 1) or (CL: term 1, # 1), and term 1 is inserted into the search term, If you enter 1, it outputs document number # 1 as the search result. On the other hand, the search engine is able to process very quickly the number of entries in the entire document or the number of claims in the claims. (Typically, the search engine returns the total number of search results first , Then take the form of first showing only the front part of the total number.)

At this time, the phrase generation engine of the present invention obtains n / terms that have passed through the morpheme analysis engine, combines the received n terms in a preset manner, and searches the combined terms together with the document number To the indexer 401-3 of the engine. (In this case, if the document number is already known by the index, it is not necessary to transmit it to the document number.) For convenience of explanation, it is assumed that a document number as a source of a combined combined term is transmitted. The method of combining is as follows: 1) a process is performed in units of one sentence, and all the terms obtained in one sentence are processed in a combination of all possible methods by 2 to 5 terms, 2) (In this case, the combination of terms is so large that the processing time is prolonged, but the most precise phrase set can be generated). ), And 3) a method of sequentially combining two to five terms in one sentence. Of course, the processing method of the entire document unit may be handled in a corresponding manner as described above, such as the processing of a paragraph unit or a specific field unit (a claim range, etc.).

For example: The phrase generation engine of the present invention receives the output, and generates 1 to 5 (preferably 2 to 3) arrays in the unit. For example, when one sentence included in a patent document is input, let the output be a b c d e f, g in order of morphemes. If the word generation unit of the word generation engine is 3 words, the word generation engine sequentially generates abc, bcd, cde, def, efg or ab, ac, ad, ae, af, ag, bc, ... fg (which is described above). At this time, the phrase generation engine transmits a phrase (= multiple terms) such as abc, bcd, ab, ac, etc. to the indexer 401-3. The indexer 401-3 stores the sentences (a plurality of terms) such as (abc, # 1), (bcd, # 1), (ab, # 1), (ac, # 1) At this time. If the fields are limited, the words (CL: abc, # 1), (CL: bcd, # 1), (CL: ab, # 1) (Fields) including the specific field names from which they are stored. Table 9 below is a conceptual table showing an example of an index.

[Table 9]

Fish Article number field abc #One D abc #One D abc #One D abc #One C abc #One C bcd #One D bcd #One C ac #One D abc #2 D abc #2 D abc #2 C

D is a description of the invention, and C is a claim.

In Table 9, the phrase abc has appeared three times in the detailed description of the invention of the document # 1, twice in the claims, and the phrase bcd is once in the detailed description of the invention of the document No. 1 And once in the claims of the patent, the phrase ac appears once in the detailed description of the invention of No. 1. On the other hand, it is shown that the phrase abc appears twice in the detailed description of the invention of item # 2, and once in the claims.

At this time, the sorting engine of the present invention receives data of the index and generates data about the number of the phrases (the plurality of terms) and the document number as the source from which the phrase is derived. The generated data is the same as or corresponding to the data generated in the above-described manner. For example, the index may be as shown in Table 10 below.

[Table 10]

Phrase ID Fish Article number field Number of times of appearance One abc #One D 3 One abc #One C 2 One abc #2 D 2 One abc #2 C One 2 bcd #One D One 2 bcd #One C One 3 ac #One D One

In this way, the representative word extraction preprocessing module of the present invention can generate categorical counting data by using the indexer 401-3 of the search engine and the index, and based on the categorical counting data, The representative phrases can be extracted from the contents of each specific field extracted from documents, predetermined document sets, specific field contents, or a predetermined document set in the same manner as in FIG.

If representative phrases are extracted for each set of representative phrase extraction target documents, the representative phrases may be stored in units of the representative phrase extraction target document set or individual document units constituting the representative phrase extraction target document set. That is, at least one representative phrase may correspond to one representative phrase extraction target document set unit or the individual document. The representative word may be corresponded to at least one of individual field units of the representative word extraction target document set or individual field units of the individual documents constituting the representative word extraction target document set.

The representative phrase extraction engine of the present invention has been described as generating the document number information together with the phrase. This is for the purpose of making it easy to extract a document set having the phrase as a representative phrase when document numbers are associated. If it is possible to easily extract a document set having the representative phrase as the phrase, the document set can be subjected to various analyzes using the document set to be analyzed. For example, for a document set with the phrase "radio frequency identification tag" as a representative phrase, 1) various analyzes such as number of applications / registration number / registration rate / market share / concentration ratio / activity rate by country / (2) analysis of the distribution of IPC or other patent classification by country, year, and applicant by depth, (3) citation of this document set by detailed document set (by applicant and patent classification) Analysis and the like, it is possible to perform various analyzes supported by the analysis module and the patent intelligence module 60 of the present invention. There is a problem in that, when a document including a specific phrase is extracted through a search engine without associating a document number with the phrase, it can not be determined whether the specific phrase is a representative word of the document.

The above-described representative word information preprocessing method is well illustrated in FIG. 105 to FIG.

105, a method for generating a combination for extracting representative phrases through the representative phrase extraction preprocessing module will be briefly described.

The representative phrase extraction preprocessing module obtains a document set including at least two patent documents (S5520), processes the entire contents of the individual patent documents extracted from the document set or the contents of various fields by a morpheme analyzer, , term or terminology) is extracted (S5530), and all possible combinations of two or more predetermined methods in accordance with a predetermined criterion are generated for the extracted information (S5540) Information is stored in the DB or transmitted to the search indexer 401-3 (S5550).

A method for generating a combination for extracting representative phrases per field / partial document set through the representative word extraction preprocessing module based on FIG. 106, and extracting a combination of combinations satisfying predetermined conditions will be briefly described.

The representative phrase extraction preprocessing module obtains a document set including at least two patent documents (S5620)

The morpheme analyzer processes all the individual patent documents extracted from the document set or contents of various fields to extract two or more terms (S5630), and extracts two or more possible combinations (S5640). The generated combination is stored in the DB including the patent document information or transmitted to the search indexer 401-3 (S5650), and the bibliography of each document constituting the document set (S5660), and obtains the number of the combinations defined in the document subset and the number of the combinations included in the document (total) set, A combination of a plurality of fields satisfying predetermined conditions is extracted (S5670)

A method for generating a combination for extracting representative phrases through the representative phrase extraction preprocessing module based on FIG. 107, and comparing the phrases according to each document set will be briefly described.

The representative phrase extraction preprocessing module obtains a plurality of document sets including at least two patent documents (S5720). Then, the representative phrase extraction preprocessing module processes the entire individual patent documents extracted from the document set or the contents of various fields by a morpheme analyzer (S5730), the extracted combination is generated (S5740), the generated combination is stored in the DB including the patent document information, or the retrieval indexer 401-3 (S5750), and the combination is compared on the basis of the plurality of document sets on a predetermined basis (S5760)

Multiple Patent Classification Relation Preprocessing Module

There are many documents in which one patent document has more than one patent classification symbol belonging to one patent classification system. All. This tendency can be attributed to a plurality of patent classification symbols for one patent document due to 1) enlargement of a patent application scope, 2) improvement of an invention, 3) tendency of technical convergence, and 4) Is becoming more and more popular. The fact that two or more patent classification marks are assigned to a patent document means that the patent document contains at least two technical views, technical constructions, technical contents, technical aspects, technical properties, technical features And the like may be included. Therefore, documents with two or more patent classification marks need to be treated differently from documents with one patent classification symbol.

The purpose of the present invention is to discover and utilize the hidden relationships among a plurality of patent classification symbols, using only documents to which at least two patent classification symbols are assigned as a material. The plurality of patent classification symbol preliminary relations preliminary processing modules find hidden relationships among the same kind of patent classification symbols based on only one of IPC / UPC / FT (that is, based on the same kind of patent classification symbol) When a patent includes two or more patent classification symbols selected from among IPC, UPC, FT and the like in the same document, And a heterogeneous plural patent classification related relationship pre-processing module for grasping the relationship between the above patent classification symbols. FIG. 60 shows the configuration of the above-mentioned plural patent classification symbol related relation preprocessing module, the configuration of the same type plural patent classification symbol relation pre-processing module, and the configuration of the different plural patent classification symbol relation preprocessing module.

Equivalent multiple patent classification symbol relationship pre-processing module

Hereinafter, the same kind of multiple patent classification symbol relation pre-processing module of the present invention will be described. Processing a plurality of patent classification symbols on a selected one of the patent classification schemes, even if a plurality of patent classification symbols of one kind are given to any one document or two or more kinds of patent classification symbols are given, The classifier relationship pre-processing module performs. The same kind of multiple patent classification symbol relationship preprocessing engine of the present invention is responsible for the processes related to the plurality of patent classification symbols, and the result of the above execution is the same as that of the present invention, ).

Theoretically, it will be explained first, followed by concrete examples. First, we propose a relationship processing method for a plurality of patent classification codes in the case where two patent classification codes are included in one document as the simplest model. Suppose that a patent document has a patent classification code A1 and a patent classification code B1. In this case, A2, A3, A4, and A3 are the parent nodes of A1, A2, and A2, respectively, in the patent classification system. ... An. Likewise, B2, B3, B4, and Bn can be assigned to B1. The top of these An and Bn are the sections. And suppose that there is a patent classification system name just like IPC on the section. When this happens, the An and Bn have at least one common node under at least a section or a patent classification scheme scheme name IPC. (Even if the sections are far apart, they share the same classification scheme called IPC. The common node among the common nodes at the lowest level in the patent classification scheme is called the lowest common patent classification symbol. At this time, Ai + 1 = Bj + 1, where Ai + 1 and Bj + 1 are the lowest common patent classification symbols of A1 and B1, respectively. If A1 and B1 are present together in the content of the given patent document, A1 and B1 are related to the patent document. (That is, (B 1, B 3,..., B j +), which are all parent nodes of the above B1, in the nature of the patent classification system. Similarly, B1 has a relationship with all the parent nodes of A1, A2, A3, ..., A (i + 1). In this case, since the relationship between A1 and Bj + 1 is the same as Bj + 1 and Ai + 1, it is not necessary to discuss the relationship between them and since Bj + 1 (= Ai + 1) Since relationships exist in parent nodes, there is no need to discuss them separately. That is, it is desirable that the lowest common parent node is not considered in the relationship calculation table.

At this time, when processing the plurality of patent document symbols, the pre-processing module for related species plural patent classification symbol may refer to the patent classification scheme (tree structure) data to produce data as shown in Table 11 below. In this case, Ai + 1 and Bj + 1 do not need cells in the actual table (only Ai and Bj are required for the relationship table) or for convenience of explanation (to show that the cells have no value) Display. That is, in the actual table, the table is composed of only Ai and Bj.

[Table 11]

A1 A2 Ai Ai + 1 B1 One One One One B2 One One One One One One One One Bj One One One One Bj + 1 = Ai + 1

That is, all the intersecting pairs (A1, B1, ..., Ai, Bj, etc.) between the A1-Ai and B1-Bj can be said to have relevance through the given patent document .

Korean Patent Application No. 10-2005-0111868 discloses H04B 7/26 and H04B 7/15 as January 2006. This is illustrated by way of example. The parent of H04B 7/26 is H04B 7/24, H04B 7/00, H04B, H04, H in turn. The parent of H04B 7/15 is, in turn, H04B 7/14, H04B 7/00. In this case, the lowest common patent classification code is H04B 7/00. Therefore, the table is shown in Table 12 below. The cell associated with H04B 7/00 has no cell value as described above.

[Table 12]

H04B 7/26 H04B 7/24 H04B 7/00 H04B 7/15 One One H04B 7/14 One One H04B 7/00

Therefore, based on Korean Patent Application No. 10-2005-0111868, nine pairs of cross-related patent classification symbols are generated. In this case, when there are two or more patent classification symbols for one document, the first one is the main patent classification symbol. Therefore, in this document, H04B 726 becomes the main patent classification symbol, Other top patent classification codes are also treated as main patent classification codes in the given patent documents. The problem of the main patent classification symbol plays an important role in relation to handling when there are three or more patent classification symbols in one document. When there are three or more patent classification symbols, the following patent classification symbols can be processed.

First, it treats all patent classification equally. When there are n patent classification symbols, the method of taking the combination of two patent classification symbols out of the n patent classification symbols mathematically becomes nC2 according to the combination theory. For each combination of the two nC2 patent classification symbols, the processing when the patent classification symbol is two is performed, and as a result, nC2 tables are obtained. In each table, information about all intersected pairs is obtained I have.

Next, the main patent classification symbol is determined, the above-mentioned main patent classification symbol is taken, the other one is taken from n-1 patent classification symbols to obtain n-1 combinations, and then n-1 tables , And each table is first treated as an equivalence of all patent classification codes. When there are n patent classification symbols, the method of taking the combination of two patent classification symbols out of the n patent classification symbols mathematically becomes nC2 according to the combination theory. As for the combination of the two nC2 patent classification symbols, the processing when two patent classification symbols are performed, and as a result, nC2 tables are obtained, and each table has information on all intersecting pairs have. At this time, weights related to the main patent classification symbol may be assigned to the crossing pair related to the main patent classification symbol. Then, n-1C2 combinations are obtained for n-1 patent classification symbols, and the same process is performed for each combination to obtain information on all intersecting pairs. At this time, weights related to the patent classification code may be assigned to the crossing pair related to the patent classification code.

Korean Patent Application No. 10-2006-0012606 discloses H04B 7/04, H04B 7/155, and H04Q 7/30 patent classification codes given as January, 2006. H04B 7/02, H04B 7/00, H04B, H04, and H are the parent nodes of H04B 7/04 and H04B 7/15, H04B 7/14 and H04B 7 / 00, H04B, H04, and H, and the parent nodes of H04Q 7/30 are H04Q 7/20, H04Q 7/00, H04Q, H04, and H, respectively. At this time, the lowest common patent classification code is H04B 7/00 for the combination of H04B 7/04 and H04B 7/155, H04 for the combination of H04B 7/155 and H04Q 7/30, H04B 7/04 and H04Q 7/15 / 30 is H04.

At this time, if all the patent classification symbols are treated equally, the following 3C2 number of tables are obtained. Table 13 is the relationship table between H04B 7/04 and H04B 7/155.

[Table 13]

H04B 7/04 H04B 7/02 H04B 7/00 H04B 7/155 One One H04B 7/15 One One H04B 7/14 One One H04B 7/00

Table 14 is the relationship table between H04B 7/04 and H04Q 7/30.

[Table 14]

H04B 7/04 H04B 7/02 H04B 7/00 H04B H04 H04Q 7/30 One One One One H04Q 7/20 One One One One H04Q 7/00 One One One One H04Q One One One One H04

Next, Table 15 is a relationship table between H04B 7/155 and H04Q 7/30.

[Table 15]

H04B 7/155 H04B 7/15 H04B 7/14 H04B 7/00 H04B H04 H04Q 7/30 One One One One One H04Q 7/20 One One One One One H04Q 7/00 One One One One One H04Q One One One One One H04

At this time, a method of assigning the following weights will be described. There are two types of weighting. There is a first way to assign a larger weight to the relevance table with the main patent classification symbol described above, and to give less weight to the relevance table between the patent classifications. For example, there is a method of assigning 1 to the relevance table with the main patent classification symbol and 0.5 to the relevance table with the patent classification symbol.

Next, the relationship tables of H04B 7/155 and H04Q 7/30 will be as shown in Table 16 below.

[Table 16]

H04B 7/155 H04B 7/15 H04B 7/14 H04B 7/00 H04B H04 H04Q 7/30 0.5 0.5 0.5 0.5 0.5 H04Q 7/20 0.5 0.5 0.5 0.5 0.5 H04Q 7/00 0.5 0.5 0.5 0.5 0.5 H04Q 0.5 0.5 0.5 0.5 0.5 H04

Secondly, there will be a way to equally weight the patent documents with multiple patent classification symbols. For example, if we have two patent classification codes, and the value of the pair of relevance of each patent classification is 1, then for pairs of patent classification codes having n patent classification codes, The value obtained by dividing nC2, which is the number of possible combinations, may be evenly distributed. That is, the weight may be 1 / (nC2). For example, when doing so, the relationship tables of H04B 7/155 and H04Q 7/30 are as shown in Table 17 below.

[Table 17]

H04B 7/155 H04B 7/15 H04B 7/14 H04B 7/00 H04B H04 H04Q 7/30 1/3 1/3 1/3 1/3 1/3 H04Q 7/20 1/3 1/3 1/3 1/3 1/3 H04Q 7/00 1/3 1/3 1/3 1/3 1/3 H04Q 1/3 1/3 1/3 1/3 1/3 H04

Third, there will be a way to assign different weightings to the pair associated with the main patent classification symbol and the pair associated with the patent classification symbol for documents with more than three patent classification symbols. For example, when there are n patent classification symbols, the number of combinations associated with the main classification is (n-1), so 75% of the total is assigned to the relevant combination with the main classifier, A weight of 0.75 * 1 / (n-1) can be given. When assigning a weight of 25% to all combinations associated with the patent classifiers, all pairs associated with patent classifiers may be weighted 0.25 * 1 / ((n-1) C2) will be. At this time, the distribution of the percentage may be adjusted to a predetermined value. For example, when n> 5, we could allocate 50% to those related to the main patent classification and 50% to those related to the patent classification, 60% to 40% when n = 4, When n = 3, you could allocate 75% to 25%. That is, when the number of IPCs is larger than 3 (sub IPC, Sub IPC), pairs of patent classification codes are formed. Therefore, the combination of (main IPC, Sub IPC) patent classification code pair and (Sub IPC, Sub IPC) We can give different weights to

When n = 3, assigning 75% to 25%, the following table can be created. Table 18 below is the relationship table between H04B 7/04 and H04B 7/155.

[Table 18]

H04B 7/04 H04B 7/02 H04B 7/00 H04B 7/155 0.75 / 2 0.75 / 2 H04B 7/15 0.75 / 2 0.75 / 2 H04B 7/14 0.75 / 2 0.75 / 2 H04Q 7/00

Table 19 below is the relationship table between H04B 7/04 and H04Q 7/30.

[Table 19]

H04B 7/04 H04B 7/02 H04B 7/00 H04B H04 H04Q 7/30 0.75 / 2 0.75 / 2 0.75 / 2 0.75 / 2 H04Q 7/20 0.75 / 2 0.75 / 2 0.75 / 2 0.75 / 2 H04Q 7/00 0.75 / 2 0.75 / 2 0.75 / 2 0.75 / 2 H04Q 0.75 / 2 0.75 / 2 0.75 / 2 0.75 / 2 H04

Next, Table 20 is a relationship table between H04B 7/155 and H04Q 7/30.

[Table 20]

H04B 7/155 H04B 7/15 H04B 7/14 H04B 7/00 H04B H04 H04Q 7/30 0.25 / 1 0.25 / 1 0.25 / 1 0.25 / 1 0.25 / 1 H04Q 7/20 0.25 / 1 0.25 / 1 0.25 / 1 0.25 / 1 0.25 / 1 H04Q 7/00 0.25 / 1 0.25 / 1 0.25 / 1 0.25 / 1 0.25 / 1 H04Q 0.25 / 1 0.25 / 1 0.25 / 1 0.25 / 1 0.25 / 1 H04

On the other hand, when the contribution of individual documents having a plurality of patent classification symbols is the same as document unit (for example, 1), the following weighting method can be considered. As an extreme example, Korean Patent Application No. 10-2005-0042032 discloses two patent classification symbols such as H04B 7/02 and H04B 7/14, and when they are in the next sibling relationship (i.e. H04B 7/02 And the H04B 7/14 parent node is the same as H04B 7/00 and its parent node is the immediate parent patent classification symbol), the relationship tables of H04B 7/02 and H04B 7/14 are only one, There is also one cell. The table for the Korean Patent Application No. 10-2005-0042032 is shown in Table 21 below.

[Table 21]

H04B 7/02 H04B 7/00 H04B 7/14 One H04B 7/00

At this time, H04B 7/02 and H04B 7/14 are strongly related to the specific patent document. On the other hand, according to Korean Patent Application No. 10-2006-0012606, three tables have 6, 16, and 20 cells for each table. At this time, when one patent classification symbol combination (H04B 7/02, H04B 7/14) is referred to, the contribution of the Korean Patent Application No. 10-2005-0042032 and the Korean Patent Application No. 10-2006- There arises a problem of how to evaluate the contribution of 0012606. The weighting process of the contribution may be performed as follows.

First, a combination of specific patent classification symbols (Ai, Bj) (for example, the above-mentioned (H04B 7/02, H04B 7/14) can be treated with the same value of contribution weight (e.g., 1).

Second, considering the number of the tables and the number of cells, contribution of each patent document to the combination (Ai, Bj) of specific patent classification symbols may be given different weights. At this time, the contribution weight for the combination (Ai, Bj) of the specific patent classification symbol can be obtained by dividing the contribution weight by the total number of all the cells existing in all tables generated in each patent document. For example, the contribution weight according to the Korean Patent Application No. 10-2005-0042032 is 1 (= 1 table, 1 cell) for (H04B 7/02, H04B 7/14) The contribution weight of the patent application No. 10-2006-0012606 is 1 / (6 + 16 + 20) (= the number of all cells in three tables).

Third, the weight of the contribution may be a value obtained by dividing the weight of the contribution by the total number of cells in the table in which the combination (Ai, Bj) of the specific patent classification symbol appears. For example, with respect to (H04B 7/02, H04B 7/14), for example, the contribution weight according to Korean Patent Application No. 10-2005-0042032 is 1 (= 1 table, And the contribution weight of Korean Patent Application No. 10-2006-0012606 is 1 / (6) (= the number of all cells in 3 tables)

Fourth, there is a method of calculating the second or third contribution weights inversely. That is, the weight of the contribution may be a value obtained by multiplying the weight of the contribution by the total number of cells in the table in which the combination (Ai, Bj) of the specific patent classification symbol appears. For example, with respect to (H04B 7/02, H04B 7/14), for example, the contribution weight according to Korean Patent Application No. 10-2005-0042032 is 1 (= 1 table, And the contribution weight of Korean Patent Application No. 10-2006-0012606 is 6 (= the number of all cells in 3 tables). This contribution weighting method will be more promising for calculating the convergence of heterogeneous technologies. In other words, when the number of cells existing in the table is small, the technology may be bundled with a strong association technique in the technology classification system itself. Therefore, the larger the size of a table in one table, This is because it is a technology apart.

The following information about the patent classification symbol combination (Ai, Bj) can be obtained by the above method.

First, it is the counting information for (Ai, Bj) (H04B 7/02, H04B 7/14). When weights are given in the present invention or in accordance with the present invention at the time of counting, it is possible to obtain the weighting-based counting information to which at least one weight is applied to the patent classification symbol combination (Ai, Bj).

Second, all patent classification combinations contain information about when the combination was created, by which document the combination was created. That is, the combination of the above-mentioned patent classification symbols (H04B 7/02, H04B 7/14) is generated by Korean Patent Application No. 10-2006-0012606 and Korean Patent Application No. 10-2005-0042032 H04B 7/15, H04B 7/24) are produced by the Korean Patent Application No. 10-2005-0111868. Therefore, a document unique number is corresponded to a patent classification symbol combination (Ai, Bj), and all the information on the document including all bibliographic references to the document having the document unique number can be corresponded to the document unique number . In all of the above tables, there are two related patent classification symbol combinations, corresponding to the patent documents from which the patent classification symbols are derived, and the bibliography of the patent documents can be corresponded to the patent documents. That is, Korean Patent Application No. 10-2005-0111868 corresponds to document identification number for all patent classification symbol combinations (Ai, Bj) existing in the relationship tables of H04B 7/155 and H04Q 7/30. In the bibliography of the document corresponding to the document identification number, the application number is 10-2006-0012606, the applicant is Samsung Electronics, the inventor is Choi and Hwang Sung Taek, and the filing date is February 09, 2006 The bibliographic information and all information related to the document corresponding to the document unique number can be corresponded. Of course, the document unique number may be an application number, and if the present patent application is filed in the US or the like, the resultant value may be mapped to the bibliography of the US application.

The correspondence between all the patent classification symbol combinations (Ai, Bj) and document numbers as described above is preferably stored in the DB together with the count value at the time of generating the patent classification symbol combination (Ai, Bj).

The above-mentioned homogeneous plural patent classification symbol relationship pre-processing module refers to the patent classification symbol system (tree structure) data for the combination of each patent classification symbol when the patent document obtained is two or more patent classification symbols, Create a predefined number of tables for each combination of symbols. At this time, the numerical value entered into each table may be a value reflecting the weight based on the predetermined reference. Then, the related-kind plural patent classification symbol preliminary processing module applies predefined contribution weight values to the combination (Ai, Bj) of the patent classification symbols extracted from the respective tables to determine the combination of patent classification symbols (Ai, Bj) (Ai, Bj) of the patent classification symbols as shown in Table 22 below.

To explain this model, suppose that there are only three patent documents: Korean Patent Application No. 10-2005-0111868, Korean Patent Application No. 10-2006-0012606, and Korean Patent Application No. 10-2005-0042032. Of course, it is possible to generate the following data for all patent classification combinations (Ai, Bj) based on all document sets by repeatedly processing all the documents obtained. , Bj) will be able to generate the following table.

[Table 22]

Patent classification symbol combination (Ai, Bj) Article number Frequency
(Simple frequency) 1 / nC2 Equal allocation (M, S), (S, S) differential allocation (H04B 7/15, H04B 7/26) 10-2005-0111868 One One One (H04B 7/15, H04B 7/24) 10-2005-0111868 One One One (H04B 7/14, H04B 7/26) 10-2005-0111868 One One One (H04B 7/14, H04B 7/24) 10-2005-0111868 One One One (H04B 7/155, H04B 7/04) 10-2006-0012606 One 1/3 0.75 / 2 (H04B 7/155, H04B 7/02) 10-2006-0012606 One 1/3 0.75 / 2 (H04B 7/15, H04B 7/04) 10-2006-0012606 One 1/3 0.75 / 2 (H04B 7/15, H04B 7/02) 10-2006-0012606 One 1/3 0.75 / 2 (H04B 7/14, H04B 7/04) 10-2006-0012606 One 1/3 0.75 / 2 (H04B 7/14, H04B 7/02) 10-2006-0012606 2 1/3 0.75 / 2 (H04B 7/14, H04B 7/02) 10-2005-0042032 2 One One One 1/3 .. One 1/3 .. (H04Q 7/00, H04B 7/00) 10-2006-0012606 One 1/3 0.25 / 1 (H04Q 7/00, H04B) 10-2006-0012606 One 1/3 0.25 / 1 (H04Q, H04B) 10-2006-0012606 One 1/3 0.25 / 1

S is the sub IPC and S is the sub IPC. If n> 2, (M, S), (M, S) , The (S, S) differential allocation is given a weight of 75% / (n-1) for (M, S) and a weight of 25% / (n-1) C2 for (S, S). When n = 2, 1 was assigned.

At this time, paying attention to the pair of patent classification symbols (H04B 7/14, H04B 7/02), it can be seen that two document numbers correspond to each other. That is, the pair (H04B 7/14, H04B 7/02) of the patent classification symbol pair appeared at the same time in two documents, and the value of frequency (simple frequency) becomes 2. 1 / nC2 equalization, and (M, S), (S, S) differential allocation.

Table 22 shows the combinations of the patent classification symbols for the three patent documents, Korean Patent Application No. 10-2005-0111868, Korean Patent Application No. 10-2006-0012606, and Korean Patent Application No. 10-2005-0042032 Ai, Bj). It will be apparent to those skilled in the art that it is possible to apply it repeatedly or recursively to all documents obtained by the method, and based on any one or more of the preset weights or contribution weights, the combination of each patent classification symbol (Ai, Bj It will be appreciated that it is possible to generate the relationship data for the &lt; RTI ID = 0.0 &gt;

As described above, the homogeneous plural patent classification symbol relation pre-processing module is a module for calculating a weighted value according to a weighting policy j for each specific situation i (1), (2), Wij), 3) whether it is (M, S) or not (S, S), and 4) other attribute information about Ai and Bj.

It will be appreciated that the method of processing the plurality of patent classification symbols by the same kind of multiple patent classification symbol relationship preprocessing module is not limited to IPC. That is, it is possible to process a plurality of patent classification symbols targeting USPC for US patent documents and FT, FI, and ECLA for Japanese patent documents exactly the same as those performed at the IPC.

Technology convergence pattern analysis module

(Ai, Bj) values stored by the pre-processing module of the same kind of multiple patent classification symbol relation with respect to the combination (Ai, Bj) of the patent classification symbols, (2) a weight number policy j 4) Whether or not at least one of the weight information (Ai, Bj) and the weight information (Ai, Bj) according to the following information , Or extract them. The following information is performed by the technology fusion pattern analysis module of the present invention.

There are essentially two pieces of information that the technology fusion pattern analysis module can process.

First, the homogeneous multiple patent classification symbol relationship pre-processing module can obtain the number of appearance of each patent classification symbol pair. For example, (H04B 7/26, H04B 7/15) appeared once in Korean patent application No. 10-2005-0111868, while (H04B 7/14, H04B 7/02) appeared in 10-2006-0012606 And 10-2005-0042032, respectively. (Ai, Bj) = (Bj, Ai) if not the main patent classification symbol or treat all patent classification symbols equally. If one of them is the main patent classification symbol (Ai, Bj) And (Bj, Ai) may be handled differently and treated the same.) Therefore, the above-mentioned homogeneous plural patent classification related relationship pre-processing module records the number of pairs of all patent classification pairs . In addition, the patent document numbers for the patent documents that are derived from the pair will be stored.

Second, the homogeneous plural patent classification symbol preliminary processing module may find the patent classification symbol most frequently combined with the specific patent classification symbol when the specific patent classification symbol is given. This may be easily achieved by counting all the pairs into which the given patent classification symbol is entered and the number of times the pair has appeared. At this time, it is possible to find a patent classification symbol that most frequently combines with the specific patent classification symbol by applying a predetermined weight or contribution weight. In addition, referring to the patent classification scheme (tree structure) You will find the patent classification code that most frequently combines in units. That is, for example, the homogeneous multiple patent classification symbol relationship pre-processing module will find the patent classification symbol most frequently associated with H04B 7/26, and it can be found at the IPC sub class, IPC main group, or IPC subgroup level You will find the patent classification code most frequently associated with H04B 7/26.

Third, the homogeneous plural patent classification symbol relation pre-processing module associates the pair of patent classification symbols with the bibliography corresponding to the patent document number through the correspondence relationship between the pair of patent classification symbols and the patent document number, Lt; / RTI &gt; The analysis result may be obtained by analyzing at least one or more combinations selected from the pair of patent classification symbols, 1) country, 2) period, 3) applicant, 4) inventor and 5) agent. For example, Samsung Electronics' patent application in Korea from 2000 to 2005 will be able to identify the most combined pairs of patent classification codes. Further, the homogeneous multiple patent classification related relationship pre-processing module may find a patent classification symbol best associated with the patent classification symbol given the patent classification symbol. In addition, given the patent classification symbol, the patent classification symbol that best combines the patent classification symbol with the patent classification symbol is referred to as 1) country, 2) period, 3) applicant, 4) inventor and 5) Can be found in the document set. For example, the homologous multiple patent classification symbol relationship preprocessing module will find the patent classification symbol for each step on the patent classification system which is most combined with H04B 7/26, and H04B 7/26, Samsung Electronics The patent application filed by the corporation from 2003 to 2005 can be found. Through this, the degree of the technical convergence shown in the patent specification for a specific subject during the specific period can be grasped.

Fourth, the homogeneous plural patent classification symbol preliminary processing module can correspond to a set of patent documents related to the combination (Ai, Bj) of the patent classification symbol for a given arbitrary combination of patent classification symbols (Ai, Bj) There will be. At this time, if the related-kind plural patent classification symbol preliminary processing module performs various analyzes on the corresponding set of patent documents, the analysis result is an analysis result on the combination (Ai, Bj) of the patent classification symbols . The set of special patent documents includes, for example, 1) a set of patent documents generated by a method of specifying a set of patent documents with a specific patent classification symbol in a specific patent classification system in a specific country DB, 2) (3) a set of patent documents by the name of a specific inventor (ie, applicant's name and inventor's name) included in the patent document of a specific applicant in a specific country DB as an inventor 4) a set of patent documents generated by a method of specifying a set of patent documents by a specific agent name in a specific country DB, 5) a set of patent documents generated by a specific country DB By a method of specifying a set of patent documents by a specific patent classification mark and a specific applicant's name 6) a set of patent documents generated by a method of specifying a set of patent documents by a specific patent classification mark on the specific patent classification system, a specific applicant's name and a specific inventor's name in a specific country DB, 7) 8) a set of patent documents generated by a method of specifying a set of patent documents by a specific applicant's name and a specific agent's name in a particular country DB, 9) a set of all patent documents in a particular country, 9) 10) a set of patent documents specified in the above-mentioned 1) to 9), 11) a set of patent documents generated by associating the specific conditions with the registration of the patent application in 1) to 9) . Of course, the same calculations can be performed on the set of documents that the user has specified or created.

In addition, the above-mentioned homogeneous plural patent classification symbol preliminary processing module is characterized in that, among the patent documents constituting the document set, at least two patent classification symbols are assigned to at least one document set specified by 1) to 11) (Ai, Bj) of the patent classification symbol can be obtained only by extracting the documents having the patent classification symbol. (Ai, Bj) of the obtained patent classification symbols, and the generated relationship data is sorted or analyzed (frequency, mode, maximum value, minimum value, number of years , The applicant, the inventor, etc.). Upon completion of the analysis as described above, it is possible to determine whether any one or more document sets specified by 1) through 11) or the technical convergence grasped from the viewpoint of the combination (Ai, Bj) of the patent classification symbol for each extracted document set subject Information can be acquired.

The above-mentioned iterative execution method performed by the homogeneous plural patent classification symbol preliminary processing module will be merely an embodiment for searching for a value generated as a result of the present execution, and those skilled in the art will be able to understand the present invention You can try different calculation methods. (For example, a table having a value (including weighting or non-granting) for each pair of patent classification symbols) for all patent classification symbols of all patent documents to which at least two patent classification symbols are assigned ) May be stored in a DB, or may be provided in a real-time generated view or a materialized view.

FIG. 103 shows an exemplary process of performing the above process. The following description will be made with reference to Fig. The similarity multiple patent classification symbol relationship preprocessing module obtains a document set including at least one patent document (S5320), and among the patent documents included in the document set, a document including a plurality of patent classification symbols having at least two patent classification symbols (S5330). In step S5330, a combination of a plurality of patent classification symbols included in the plurality of patent document documents is generated in accordance with predetermined criteria regarding classification processing of the main patent classification symbol and the patent classification symbol (S5340) Information on the parent node in the patent classification symbol system is obtained for each patent classification symbol included in the combination (S5350), the lowest common patent classification symbol is obtained for the parent node (S5360) A pair of each parent node patent classification code including itself up to the patent classification symbol is generated (S5370), and each pair Stores the weighting applied uniformly or pre-set criteria, and (S5380), and generates the statistical values, parameters or values calculated according to predetermined criteria to the destination for each pair of the stored (S5390)

Next, a method for preprocessing and comparing a plurality of patent classification symbols with respect to at least two document sets through the homogeneous plural patent classification symbol preliminary processing module will be described with reference to FIG.

The same kind of multiple patent classification symbol relationship pre-processing module obtains at least two sets of documents including at least one patent document (S5410), and, among the patent documents included in the document set for each document set, two patent classification symbols (S5430), and generates a combination of a plurality of patent classification symbols included in the plurality of patent document documents according to predetermined criteria regarding classification processing of the main patent classification symbol and the patent classification symbol (S5440), information on the parent node in the patent classification symbol system is acquired for each patent classification symbol included in the combination for each combination (S5450), the lowest common patent classification symbol is obtained for the parent node (S5460), a pair of each parent node patent classification symbol including itself up to the lowest common patent classification symbol (S5470). Each of the pairs is weighted by applying an equal or preset reference (S5480), and a statistical value, a parameter or a calculation value according to a preset reference is generated (S5490), and compares the statistical values, parameters, or calculated values on the document set (S5495).

Figures 115 to 117 illustrate an exemplary embodiment of a plurality of patent classification symbol analysis methods.

115 is an exemplary method for obtaining patent classification codes and presenting analytical information on the most-recent fusion patent classification symbols based on the acquired patent classification codes.

The same kind multiple patent classification symbol preliminary processing module receives at least one or more patent classification symbols (S6520), obtains the most-fused patent classification symbol information for the acquired patent classification symbols (S6530) At least one step is displayed in accordance with the patent classification scheme, or at least one analysis is performed on the set of patent documents corresponding to the best-fit patent classification symbol in operation S6540.

116 is an exemplary method for acquiring a document set, extracting a most-favored patent classification symbol from the acquired document set, and presenting analytical information on the least-fused patent classification symbols based on the extracted patent classification symbol .

The similarity multiple patent classification symbol relation pre-processing module obtains at least one document set (S6620), obtains a patent classification symbol for individual documents constituting the document set (S6630), obtains the patent classification symbol obtained for each document (S6640), the most-convergent patent classification symbol information is obtained for each individual patent classification symbol ranked (S6650), and the patent classification symbol system for the least-convergent patent classification symbol is obtained (S6660), or at least one analysis is performed on the set of patent documents corresponding to the least-fused patent classification symbol.

FIG. 117 is a view for explaining the method of extracting the most-fused patent classification symbols based on the patent classification code extracted from the document unit in the acquired document set, integrating the extracted most-used patent classification symbols, Which is an exemplary method.

In step S670, the pre-processing module acquires at least one document set (S670), acquires a patent classification symbol for each document constituting the document set (S670), obtains the patent classification symbol obtained for each document (S670), the most-convergent patent classification symbol information obtained for each document is integrated (S670), and the integrated most-favored fusion classification symbol is reflected at least (S670) at least one step or at least one analysis for the set of patent documents corresponding to the integrated best-fit patent classification code is performed.

The above-mentioned homogeneous multiple patent classification symbol relation pre-processing module may be applied to a patent classification (PC) symbol pair (PC1, PC2) included in each document without applying the subpopulation symbol included in the feature of the present invention. ..., PCi, PCj, ..., PCn. That is, in the above description, it is possible to construct a pair of patent classification symbols for only (PCi, PCj) shown in the document without considering the upper patent classification symbol on the given patent classification scheme for a given PCi. At this time, since PC1 is the main patent classification symbol, the main &lt; RTI ID = 0.0 &gt; The weighting between sub can be applied as it is.

Heterogeneous multiple patent classification symbol relationship pre-processing module

The homogeneous multiple patent classification related pre-processing module described the processing of the same kind of patent classification code. Next, the heterogeneous plural patent classification symbol relation pre-processing module will be described. The heterogeneous plural patent classification related relationship pre-processing module is completely similar to that of the homogeneous plural patent classification related pre-processing module. In the United States, patent documents contain almost all IPC and USPC combinations in one patent document, and both IPC and FT are granted to patent documents issued in Japan. One patent document on IPC1, ... IPCn, USPC1, ... , And USPCm. In this case, in view of the heterogeneous plural patent classification relation processing, the IPC1, ... IPCn and USPC1, ... , There are four kinds of relationship combinations between USPCm. First, in relation to IPC1 and USPC1, they are a pair of main IPC and main USPC, respectively. And, IPC1 and USPC2, ... , A pair of USPCm, which is a pair of a main IPC and a sub USPC. Then, IPC2, ... A pair of IPCn and USPC1, which is a pair of sub IPC and main USPC. Then, IPC2, ... Each IPCn and USPC2, ... , And USPCm, which is a pair of sub IPC and sub USPC.

The above-mentioned heterogeneous plural patent classification symbol relationship pre-processing module can perform the same processing as that performed by the homogeneous plural patent classification symbol relationship pre-processing module among different heterogeneous patent classification symbols under each type for each four kinds. At this time, because the heterogeneous patent classification symbol is targeted, since there is no minimum common patent classification symbol, it can be limited to a predetermined value. That is, the upper patent classification code can be defined for a given IPC and a given USPC pair, targeting only the subclass in the case of IPC and the class in the USPC. That is, the above-mentioned homogeneous plural patent classification symbol relation preprocessing module generates the above relationship table for a given patent classification symbol combination (Ai, Bj) (where Ai and Bj are patent classification codes of the same system) As described above, the heterogeneous plural patent classification related relationship preprocessing module generates the above relationship table for the patent classification symbol combination (Ai, Bj) (where Ai and Bj are patent classification codes of different systems) . A weight or a contribution weight may be given to the generated relationship table. In addition, it is possible to correspond the patent document number to each of the heterogeneous patent classification symbol combinations (Ai, Bj) (where Ai and Bj are patent classification codes of different systems), and the bibliography It will also be appreciated that counting and calculation may be performed to reflect the matter. In addition, for a particular set of documents, a heterogeneous patent classification combination (Ai, Bj) may be obtained and an equivalent analysis of the homogeneous multiple patent classification related preprocessing module for these combinations may be performed.

Statistical preprocessing module by patent classification symbol

Multistage of patent classification codes

The statistical preprocessing module for each patent classification symbol according to the present invention performs preprocessing for finding a predetermined statistical value for at least one patent database of at least one country with respect to a patent classification symbol on at least one patent classification system.

FIG. 61 illustrates the structure of the statistical preprocessing module for each patent classification symbol. The statistical preprocessing module for each patent classification symbol includes a statistical preprocessing engine 3210 for each patent classification symbol for preprocessing statistical values for each patent classification symbol, a view storing statistical values for each preprocessed patent classification symbol, ) Or a preprocessed patent classification symbol-specific statistical DB 3230 as an information organization.

The patent classification symbol handled by the statistical preprocessing module according to the patent classification symbol is at least one of IPC, USPC, FT, FI, and ECLA. In the statistical preprocessing module for each patent classification symbol, the patent database of each country includes the patent database issued by Korea, USA, Japan, Europe (EPO), and the patent database issued by the patent office of other countries Do. Each country has a common IPC and may have a special classification system (USPC or UPC for US Patent Office, FT, FI for European Patent Office, ECLA for European Patent Office) in each country. It is desirable that the pre-set statistical value for a patent database of a country having two or more patent classification systems is independently performed for each of the above two or more classification systems. That is, in the case of the United States, it is necessary to preprocess IPC and USPC respectively. In addition, each patent classification system has a level for each system. In the case of IPC, there are section, subsection, class, subclass, group, and subgroup. In subcategories below, the hierarchical structure is expressed by the number of dots in the title information corresponding to the patent classification symbol. An example of the hierarchical structure is described in detail in the patent classification pre-processing module 301-3-1 or 3500. FIG. Therefore, the level of the patent classification symbol in the present invention can be divided into IPC section, class, subclass, main group, 1 dot subgroup, 2 dot subgroup, ... n dot subgroup. The level can identify the upper and lower hierarchical structures in the classification symbol system up to the main group, but from the subgroup, the dot information contained in the title must be utilized to confirm the hierarchical structure of the upper and lower levels. It is from the subclass in the case of USPC to check the hierarchical structure as the number of dots as above, and from FT in the case of FT after all 7 digits (from "theme code + two letters of English alphabet") to the whole FT symbol.

An example of a method of preprocessing statistical values for each patent classification symbol by the statistical preprocessing module for each patent classification symbol is shown in FIG. The statistical preprocessing module for each patent classification symbol may be used to readjust the patent classification symbol by using the tree structure of the patent classification symbol so that the sub classification code of the given patent classification symbol is automatically included in the specific patent classification system in the first country (S5020). Then, the retrieval expression or the retrieval query including the patent classification symbol is obtained (S5030), and the retrieved expression or the retrieval query is converted into the re-adjusted patent classification symbol for the given patent classification symbol included in the retrieval expression or the retrieval query S5040), the search engine or the DBMS 201 searches for a retrieval expression or a retrieval query including the re-adjusted patent classification symbol (S5050), obtains and quantifies the patent document information as the retrieval result (S5060) , An applicant, an inventor, or a representative of the applicant, on a predetermined time unit or an integrated time unit with respect to the search result The statistical value, and performs a function of calculating the ranking information based on the parameters or calculations.

In addition, the statistical preprocessing module for each patent classification symbol may automatically generate a statistical value, a parameter or a calculated value by automatically including a sub-patent classification symbol for a given patent classification symbol, and an exemplary method thereof is shown in FIG. 101 .

The statistical preprocessing module for each patent classification symbol obtains a search formula including a patent classification symbol (S5120). The patent classification information includes patent information related to the sub-patent classification symbol for the patent classification symbol included in the search formula, (S5130), and processes the statistical values, parameters, and calculation values of the citation information and the cited information included in the patent document (S5140).

In addition, the statistical preprocessing module for each patent classification symbol generates a statistical value, a parameter value, or a calculation value according to a patent classification symbol from the viewpoint of including a sub-patent classification symbol for a citation or a cited information for a document subset of a specific document set An exemplary method for this is shown in FIG.

The statistical preprocessing module for each patent classification symbol obtains at least one document set (S5220), extracts at least one patent technical classification symbol from the applicant, the inventor, and the proxy for all documents included in the document set (S5230 ), A document subset is generated by applying one or more of the extracted criteria to the document (S5240), and for all the documents constituting the regular document subset, the first-party document information cited by the document is obtained (S5250). Then, for all the documents constituting the document subset, the application document information after citing the document is obtained (S5260). Then, the obtained document and / or the document Quot ;, a related statistic value for a citation, a parameter or a calculation value (S5270)

Hereinafter, the statistical preprocessing module according to the patent classification symbol of the present invention will be described in more detail from the viewpoint of the statistical preprocessing engine for each patent classification symbol.

Options

When the statistical preprocessing module according to the patent classification symbol processes the preset statistical value, the following options are considered, and each option can be cross-selected.

First, it is a matter of whether information on the subclassification symbol of the patent classification symbol is automatically included when each patent classification symbol is given. This is especially true if you need to check the hierarchy from the number of dots in the title information. For example, using the extension to H, H04, H04B, H04B 7/00 levels, it is easy to find patent documents containing patent classification symbols at the bottom of a given level of IPC, but at the subgroup level, Only the sub-classification search method can be used to find patent documents containing patent classification codes belonging to the sub-category. For example, if you have an IPC of H04B 7/15 and you want to generate statistical values without subcategorizations, you can obtain a patent document containing H04B 7/15 from each country's patent documents. However, when generating statistical values, including subclassifiers, H04B 7/15 and H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19 under the IPC hierarchy , H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, H04B 7/216. This also applies to the USPC, FT, and the like, and a common description will be omitted.

When processing the statistical value including the sub classifier automatically, the range search of the present invention (both using the search engine or using the DBMS 201) may be utilized. In order to achieve this object, when processing a given patent classification symbol, not all of the range search, all the patent classification symbols below the patent classification symbol are stored, and when processing the statistical value for the patent classification symbol, It is possible to process statistical values for all patented classifiers in the subordinate, rather than only statistical processing for symbols. H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B at the bottom of H04B 7/15, for example, 7/204, H04B 7/208, H04B 7/212 and H04B 7/216 from the Patent Classification Hierarchy Information, then H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, H04B 7/216, H04B 7/155, H04B 7/165, H04B 7/17, H04B 7/185, H04B 7/19, H04B 7/195, H04B 7/195, H04B 7/195, H04B 7/195, H04B 7/195, H04B 7/195, H04B 7/204, H04B 7/208, H04B 7/212, and H04B 7/216 may be adopted. This is true of other patent classification schemes such as USPC and FT.

When there is a patent classification code of a given level, it is preferable to automatically include the sub-patent classification codes of the patent classification symbol rather than generating the predetermined statistical value for a document containing the patent classification code, . &Lt; / RTI &gt;

Second, if there are multiple patent classification codes, how to handle them is a problem. In this case, 1) the main patent classification code (usually the first one is the main patent classification code) is used as the statistical processing target, 2) the main patent classification code and the sub patent classification code (Eg, when there are n + 1 patent classification codes, the main patent classification code is 50% or more) , And the rest of the patent classification code is weighted by 50% * 1 / n). Of the above 1) to 3), 1) and 3) are preferable, and 3) among them is more preferable.

Third, it is a processing method when there are a plurality of applicants and inventors. In this case, 1) a method of giving a weight of 100% to each person who made a sale or an inventor, 2) a method of balancing the number of applicants or inventors (100% * 1 / n for each inventor when there are n inventors) . In the case of the applicant, it would be desirable for the inventor to treat it as 1 / n to give a weight of 100% for each applicant in terms of implementation independence. In the case of an agent, the processing method of the applicant or the inventor as described above may be adopted, and it is preferable that the agent is 100%.

The statistical preprocessing module for each patent classification symbol may count the number of its own sub-classifications on the patent classification system for a given patent classification symbol and count the number of all its sub-patent classification symbols It will be possible.

The statistical preprocessing module for each patent classification symbol may calculate the following information on the basis of the application standard and / or registration standard for each predetermined period of time, with respect to the specified patent DB for the given patent classification symbol. The reason for this calculation is that a set of patent documents satisfying the conditions such as the patent classification symbol, the designated country, the set time, the application or the registration is determined. By counting the determined set of patent documents, various parameters such as the following can be calculated.

For example, the number of applications, the number of applications, the rate of increase in the number of applications, and the acceleration of the number of applications can be calculated by the year or quarter. The number of applications per year may be calculated in such a manner that the total number of patent documents filed in a particular year is obtained. The calculation of the growth rate is {(current period number - past period number) / past period number} * 100%, and the increase rate is {(current period number - past period number) / time interval} The rate of change with respect to the time interval of the increase rate.

In addition, the number of applicants per year may be obtained (in this case, in the case of a joint application, the above options may be applied), the number of applicants can be calculated when the number of applicants is calculated, The rate of increase in embroidery and average number of applications will also be calculated. The number of inventors, the number of inventors, the average number of inventors, and the average number of inventors can be calculated based on the number of applicants per year by the year.

In addition, since the number of patent claims can be calculated for each individual application, the number of claims for patent application, the number of claims for patent application, the average number of patent applications per application, and the average patent application claim increase rate can be calculated. In this case, if the claim is divided into independent and dependent clauses, the number of independent clauses and the number of dependent clauses can be calculated.

In addition, since the number of patent classification symbols can be calculated for each individual application, it will be possible to calculate the rate of increase of the Patent Classifier Lake and Patent Classifier Lake by year, and the average Patent Classifier Lake and the average Patent Classifier Lake Growth Rate per application can be calculated will be. In the case of a patent database in which two or more patent classification codes are used, such as the United States or Japan, it may be calculated for each patent classification code. When there is family information about the application for each individual application, the number of families by year, the number of milliseconds to increase, the average number of family countries per application, and the average family number increase rate can also be calculated based on the family information.

The above is calculated based on the application standard, and the same parameters may be calculated on the basis of registration. The parameters that can be calculated based on the registration criteria include the number of registrations, the number of registrations, the number of registrations, the number of registrants, the number of registrants, the number of registrants, the average number of registrants, the average number of registrants, Patent Classifier Lake Growth Rate, Average Patent Classifier Lake, Average Patent Classifier Lake Growth Rate, Number of Families, Number of Mills, Number of Claims, Number of Registered Claims, Growth rate, average number of family countries, and the average number of family countries.

The above parameters are exemplary, and various parameter values may also be calculated using various numerical information included in the application or registered patent information and the pre-counted values of the present invention. The parameter values are essentially a count value, a change rate and a change value of the count value, a statistical amount (average, standard deviation) of the count value, and a function value (e.g., a registration rate) for two or more count values.

Discovery of quantity subject

The statistical preprocessing module for each patent classification symbol may calculate the following information on the basis of the application standard and / or registration standard for each predetermined period of time, with respect to the specified patent DB for the given patent classification symbol. The reason for the above calculation is that a set of patent documents satisfying the conditions such as the patent classification symbol, the designated country, the set time, the filing or the registration is specified. The counting of the specified set of patent documents makes it possible to calculate various parameters as follows.

For example, by counting the set of specified patent documents, the inventor or agent who is the most applicant of the year or the year integrated can be extracted and ranked by each applicant, inventor, or agent.

In addition, it is possible to calculate the concentration, activity index, etc. for the applicants, inventors, or agents extracted from the above, and the ranking of the applicant, inventor, or agent based on the degree of concentration as a result of the calculation , And the ranking of the applicant, inventor, or agent based on the activity level may be obtained. For reference, the concentration based on the number of applications will be calculated as {(100%)} (all applications / specific applications for a particular subject for a particular time period). In addition, the activity level is defined as [({(all applications for a specific subject's specific patent classification for a particular time period / all applications for a particular subject during a specified period)} / { / All applications for a specific period)} * 100%]. 1) All applications for specific patent classification codes of a specific subject during a specific period, 2) All applications of a specific subject during a specific period, 3) All of the applications for the classifier, and 4) the number of each of the applications for a particular period of time, so that the above concentration or activity can be calculated. Meanwhile, the concentration or activity is only an example of a patent index, and a defined function that takes as input values numerical values indirectly obtainable directly or through pre-calculation to the search engine or the DBMS 201 If so, it will be able to calculate the value of the function and calculate the ranking based on the function value of the applicant, inventor, or agent through the calculated function value. In the function of the present invention, all patent analysis indicators defined as values obtainable by the search engine or the DBMS 201 on the basis of the patent DB among all the conventional patent analysis indicators that can be defined as the patent analysis index .

In addition, in the above description, the rankings of applicants, inventors, or agents are calculated based on the number of applications, but the rankings of applicants, inventors, or agents are determined for the patent indexes or functions based on other parameters such as the number of applications, . The parameters include the number of claims to be filed, the number of claims to be filed, the average number of claims per patent application, the average number of patent claims per patent application, the patent classifier lakes, the patent classifier lake growth rate, There may be an average patented classifier rate per filing, the number of families, the number of millions, the average number of family countries per application, and the average number of family members per application.

Although various patent indices or functions have been described with reference to the application, it will be obvious that they can also be defined and calculated based on registration. This is because the limitation of the application or registration is a matter of specifying or determining the set of patent documents. Therefore, the patent document set may be specified or determined only for the application for which the examination request is made at the application stage among the applications. In addition, in the case of introducing the utility model system, whether or not to include the utility model registration application or the utility model registration in each application or registration is also a problem concerning the specification or determination of the set of patent documents, May be variously determined at the level of those skilled in the art.

Citation related

The statistical preprocessing module for each patent classification symbol is a patent database (for example, a patent database established by the United States Patent and Trademark Office patent document) containing citation information for a given patent classification symbol, , The citation related information may be generated based on the application standard and / or the registration standard. The reason for the above calculation is that a set of patent documents satisfying the conditions such as the patent classification symbol, the designated country, the set time, the filing or the registration is specified. For example, when USPC = 002/456 (body cover among GUARD OR PROTECTOR among MISCELLANEOUS among Apparrel) is given, a set of patent documents containing USPC = 002/456 as a patent classification code of own patent document . The counting of the specified set of patent documents makes it possible to calculate various parameters as follows.

In the bibliography of US patent documents, there is citation information such as patent of other person quoted by his / her own document. In other words, when document B is citing documents A and a, documents A and a are cited documents, ie patents are patents, and documents B are patents citing documents A and a. In this case, the applicant (including the asignee), the inventor, the agent and the patent classification code related to the document B become the patent applicant, the patentee, the patent agent, the patent agent, and the patent classification code, respectively. The inventor, the agent, and the patent classification code shall be patent categorizations for the patentee, the patentee, the patent agent, and the patentee, respectively.

When the B document is citing the A document and the a document, the bibliography of the document B contains the numbers of the A document and the a document. When the B document has a relationship citing the A document and the a document, mapping from A-> B, a-> B and B-> A, B-> a are computationally cited , And the relationship of A <-> B, a <-> B can be established through the mapping relation. Therefore, a document that B is citing can be easily dataized in B's bibliography, and through the mapping, a set of documents to which A is being cited and a set of documents to which a is cited can easily be specified have. That is, the set of documents for which the A document is cited and the document for which the a document is cited will include the B document.

The process may be performed on all documents that have been confirmed or specified. That is, you can set mapping relationships between all patent documents when they are processed one by one for all specified patent documents, or in other computational ways. When processed one by one, it will be obvious that the document to be processed is placed in the position of document B. In this case, if both the A and the a documents are included in the overall set of US patent DB, the information related to the document A and the document a can be recorded in the information about the document B, Even in the case of a managed document, the information related to the document in the patent database of the country may be recorded in the information related to the document.

If a set of patent documents containing citation information (a set of documents with B) is specified, the statistical preprocessing module for each patent classification symbol is a patent document number (refer to document A The number of the document a, and the number of the document a) can be collected. The collected patent document number includes the corresponding patent document, and it is obvious that the patent document for citing includes the bibliography. Accordingly, a set of patent documents for citing documents cited by all patent documents constituting the specified patent document set is also specified at a specific calculation time reference. Therefore, the specified set of patent documents for citing can be analyzed, counted, or calculated. At this time, redundant counting is, of course, allowed. That is, when a cited document is cited by a plurality of documents of a set of patent documents containing the citation information, the cited document must have a weight or a count value as many times as the cited document. That is, it is valid if each cited document number is deemed to be multiplied by the weight of the number of citations. That is, when document A is cited five times and document a is cited three times, when calculating the following parameter values such as the number of cited applications, a weight of 5 for the A document and a weight of 3 for the a document . Based on a set of patent documents containing citation information, all calculated values for the cited patent document set may be referred to as "quoted ". When the duplicate counting is performed, the rankings of cited document numbers quoted the greatest number of times will be calculated.

The statistical preprocessing module according to the patent classification symbol may calculate the following information about a set of patent documents for the cited document (document A, a document set with document a) or a specific set of patent documents, and The number of cited applications, the number of cited applications, the rate of increase in the number of cited applications (these can be known by the number of documents constituting the set of patent documents for cited persons) can be known, and the increase in the number of cited applications, The number of patent applicants, the number of patent applications, the average number of patent applications, the average number of patent applications, the number of patent inventors, the number of patent inventors, the average number of inventors, the number of patent inventors, the number of patent applications, Number of patent claims, number of filed patent applications, number of claims filed, patent categorizer lake, patent categorizer lake growth rate, citation average patent classifier lake, The growth rate of the patent classifier lake could also be calculated. The number of cited applications increased, the number of cited applications increased, the number of cited applications increased, the number of applications cited, the number of applications cited, the average number of applications cited, the average number of applications increased, the number of cited inventors, , The average number of inventors cited, the average number of inventors increased, the number of claims filed, the number of claims filed, the number of patent applications filed, the average patent filings filed, the number of filed patent claims, the patented classifier lake, The method of counting the classifier lakes, the average patent classifier lake growth rate, and the like is the same as described in the above parameter calculation method. In other words, with respect to a set of patent documents for citing, the number of applications, the number of applications, the rate of increase in the number of applications, the acceleration of the number of applications, the number of applicants, the number of applicants, the average number of applicants, Patent Classifier Lake Growth Rate, Average Patent Classifier Lake, Average Patent Classifier Lake Growth Rate, Number of Average Inventors, Average Number of Inventors, Number of Claims, Number of Claims, The rate of increase in the number of cited applications, the rate of increase in the number of cited applications, the rate of increase in the number of cited applicants, the rate of increase in the number of cited applicants, the number of cited average applications, the rate of increase in the number of cited average applications, The number of inventors, the number of inventors, the average number of inventors, the average number of inventors, the number of patent applications, the number of patent applications, the number of patent applications, , The rate of increase in the number of patent application claims, the rate of patent categorizer lake, the rate of patent categorizer lake increase, the average patent classifier lake, and the average patent classifier lake rate of citation.

Similarly, the statistical preprocessing module for each patent classification symbol may calculate the ranking of the applicant, the inventor, the agent, or the patent classification symbol for each of the parameters for the patent document set for the cited person. That is, it is possible to calculate the ranking of the applicant for the highest number of applicants based on the number of applications, and the ranking for each inventor by the highest number of applicants. At this time, when calculating the rankings according to the patent classification symbols, the ranking of the most cited patent classification symbols may be calculated for each stage of each patent classification system using the sub-patent classification scheme of the present invention. At this time, at least one or more patent classification symbols included in the patent document for citing included in the set of cited patent documents will be included, and all the upper patent classification codes (in the patent classification scheme) It would be reasonable to assume that it was quoted. Therefore, it would be desirable for this view to be included in the calculation of the quotation of the most cited patent classification symbols for each step. When the US patent documents included in the set of patent documents for citing are targeted, the patent classification codes may be both USPC and IPC.

The statistical preprocessing module for each patent classification symbol can acquire applicant information on a patent document set (document set with B) containing the quotation information, and it is possible to obtain applicant information A set of documents (a set of patent documents per applicant of any one of the above-mentioned applicants will include the document B), and the set of patent documents for the cited persons is also specified for the set of patent documents for each applicant And it will be possible to calculate various parameters such as the number of cited applications for the specified set of cited patent documents. At this time, the applicant's ranking can be calculated for each parameter. For example, it is possible to extract applicants from a set of patent documents identified by USPC = 002/456 (body cover among GUARD OR PROTECTOR among MISCELLANEOUS among Apparrel, by year), and to generate a document set for each extracted applicant (For example, US Patent No. 06401262, assignee Benetton Group SpA, US Cl. 2/456; 2/411), the document set corresponding to the above Benetton Group SpA, USPC = 002 / The statistical preprocessing module for each patent classification symbol can generate the citation document set in the same manner for the document set generated for each applicant, Can be calculated.

From the above mapping point of view, a set of documents citing each patent document included in the specified set of patent documents, that is, a set of documents cited, may be generated. For example, when a B document is cited by a C document and a c document, it can be seen that the B document is cited by the C document and the c document. To find these C and c documents, enter the number of the B document in the citation document number field of the search engine. Then, the C document containing the B document in the citation information appears as the search result. Of course, it is possible to obtain a C document or the like as a search result by querying a query such as select by designating a quotation number field for the number of the B document in the DBMS 201. If the mapping relations of B-> C and B-> c are summarized as data, it is possible to easily specify a document set including the C document citing the B document.

Therefore, the statistical preprocessing module for each patent classification symbol may be used for a set of patent documents citing the individual patent documents of the specified patent document set for a specified set of patent documents, In the case of B, the documents containing the quoted C are the information for reference).

If a set of patent documents (a set of documents with B) is specified, the statistical preprocessing module for each patent classification symbol queries the search engine or the DBMS 201 to obtain a patent document citing each patent document of the patent document set The number (the number of document C, the number of document c). Since it is obvious that the bibliographic references correspond to the collected patent document numbers, the set of patent documents citing all the patent documents constituting the specified patent document set is also specified at a specific calculation time reference. Therefore, the specified set of cited patent documents can be analyzed, counted, or calculated.

At this time, it is preferable that duplicate counting is allowed. If both the B document and the B document are cited by the C document, the weight of the document C should be two. That is, the C document is included in the search result even if the B document is searched.

Based on the specified set of patent documents, all calculated values for the cited patent document set may be referred to as "for citation. &Quot; When the duplicate counting is performed, the ranking of the citation document number that cites the most patent documents belonging to the specified set of patent documents will be calculated.

The statistical preprocessing module according to the patent classification symbol may calculate the above information on the cited patent document set (document C, document c with document c).

The statistical preprocessing module for each patent classification symbol may similarly calculate the ranking of the applicant, inventor, agent, or patent classification symbol for each of the above-mentioned parameters for the cited patent document set. That is, it is possible to calculate the ranking for the most cited applicant and the ranking for the most cited inventor based on the number of applications. At this time, when calculating the rankings according to the patent classification symbols, the ranking of the patent classification symbols for each of the plurality of patent classification schemes can be calculated by using the sub-patent classification scheme of the present invention. At this time, at least one or more patent classification symbols included in the cited patent documents included in the cited patent document set will be included, and each of the upper patent classification codes (in the patent classification scheme) It would be reasonable to assume that Therefore, it would be desirable for this view to be included in the calculation of the quotation of the most cited patent classification symbols for each step. When the US patent documents included in the cited patent document set are targeted, the patent classification code may be both USPC and IPC.

At this time, the statistical preprocessing module according to the patent classification symbol can analyze the specified patent document set (the document set having B and b) itself. Even at this time, it is preferable that duplicate counting is allowed, and weights are assigned to the duplicated number of times. For example, when document B is cited five times and document b is cited three times, when calculating the following parameter values, such as the number of applications for culling, for all the specified set of patent documents, , And a weight of 3, and a weight of 3. The statistical preprocessing module for each patent classification symbol may calculate the ranking of the most cited patent documents on the specified set of patent documents by reflecting the weight. Further, the statistical preprocessing module according to the patent classification symbol is used for the number of cited patent applications (C is cited by C) on a quantitative basis, with respect to the specified set of patent documents (documents B and B) ) Can be used to count or compute values such as.

The statistical preprocessing module for each patent classification symbol may acquire the applicant information on the specified set of patent documents (set of documents with B), and the set of patent documents for each applicant (any one of the above-mentioned best applicants The set of patent documents for each applicant will be included in the set of patent documents for each applicant), and the set of patent documents for the citation It is possible to calculate various parameters such as the above-mentioned number of cited applications for the set. At this time, the applicant's ranking can be calculated for each parameter. For example, it is possible to extract applicants from a set of patent documents identified by USPC = 002/456 (body cover among GUARD OR PROTECTOR among MISCELLANEOUS among Apparrel, by year), and to generate a document set for each extracted applicant (For example, US Patent No. 06401262, assignee Benetton Group SpA, US Cl. 2/456; 2/411), the document set corresponding to the above Benetton Group SpA, USPC = 002 / The statistical preprocessing module for each patent classification symbol can generate the citation document set in the same manner with respect to the document set generated for each applicant, Parameters can be calculated.

The method for calculating the various parameters by the statistical preprocessing module according to the patent classification symbol according to the present invention is characterized in that calculation values are generated for each defined or set patent document set. Therefore, in the view of the statistical preprocessing module according to the patent classification symbol, it is important that the set of the specified patent document inputted to the statistical preprocessing module according to the fraud patent classification symbol is important, and the property and the size of the patent document set do not matter , Citation information must be included in the specified set of patent documents in order to perform calculations related to citations.)

In addition, the statistical preprocessing module for each patent classification symbol of the present invention may calculate at least one of the above parameters for a set of patent documents designated by the user using the statistical preprocessing module for each patent classification symbol.

Patent Information Processing Base Module (40)

Next, the patent information processing basic module 40 of the present invention will be described. The patent information processing basic module 40 includes a search engine module 2) an operation execution result table generation module 402 for multidimensional analysis, 3) an analysis module, 4) a monitoring module 403, and 5) Module 6) a directory creation module 405, 7) a reporting module 406, 8) a simple analysis module 407, and the like.

Heterogeneous multiple patent classification symbol relationship pre-processing module

The above-mentioned homogeneous plural patent classification related pre-processing module described the processing of the same type of patent classification symbol. Next, the heterogeneous plural patent classification symbol relation pre-processing module will be described. The heterogeneous plural patent classification related relationship pre-processing module is completely similar to that of the homogeneous plural patent classification related pre-processing module. In the United States, patent documents include almost all IPCs and USPCs combined in a single patent document, and IPCs and FTs are assigned to patent documents issued in Japan at the same time. Suppose IPC1, IPCn, USPC1, ... USPCm are assigned to one patent document.

In this case, in view of the heterogeneous plural patent classification related processing, the IPC1, IPCn,

There are four kinds of relationship combinations between USPC1, ..., and USPCm. First,

In relation to USPC1, they are each a pair of main IPC and main USPC. Then,

A pair of USPC2, ... USPCm, which is a pair of a main IPC and a sub USPC. Then, in the pair of IPC2, IPCn and USPC1, this is a pair of sub IPC and main USPC. Then, pairs of IPC2, ..., IPCn and USPC2, ..., USPCm, respectively, are a pair of sub IPC and sub USPC.

The above-mentioned heterogeneous plural patent classification symbol relationship pre-processing module can perform the same processing as that performed by the homogeneous plural patent classification symbol relationship pre-processing module among different heterogeneous patent classification symbols under each type for each four kinds. At this time, because the heterogeneous patent classification symbol is targeted, since there is no minimum common patent classification symbol, it can be limited to a predetermined value. That is, the upper patent classification code can be defined for a given IPC and a given USPC pair, targeting only the subclass in the case of IPC and the class in the USPC. That is, the above-mentioned homogeneous plural patent classification symbol relation preprocessing module generates the above relationship table for a given patent classification symbol combination (Ai, Bj) (where Ai and Bj are patent classification codes of the same system) As described above, the heterogeneous plural patent classification related relationship preprocessing module generates the above relationship table for the patent classification symbol combination (Ai, Bj) (where Ai and Bj are patent classification codes of different systems) . A weight or a contribution weight may be given to the generated relationship table. In addition, it is possible to correspond the patent document number to each of the heterogeneous patent classification symbol combinations (Ai, Bj) (where Ai and Bj are patent classification codes of different systems), and the bibliography It will also be appreciated that counting and calculation may be performed to reflect the matter. In addition, for a particular set of documents, a heterogeneous patent classification combination (Ai, Bj) may be obtained and an equivalent analysis of the homogeneous multiple patent classification related preprocessing module for these combinations may be performed.

Patent Classification Search Module (401)

Next, the patent classification code search module 401 will be described. The search of a patent classification symbol refers to outputting, as a search result, a patent classification symbol including the technical keyword as title information when a technical keyword is inputted as a search term.

Next, the characteristic patent classification search of the present invention will be described with the following examples.

Section: H Electricity

Class: H01 Basic electrical element

Subclass: H01F Magnets

Main group: H01F 1/00 Magnet or magnetic body characterized by magnetic material

1-dot subgroup: H01F 1/01 * Inorganic materials

2-dot subgroup: H01F 1/03 ** characterized by coercivity

3-dot subgroup: H01F 1/032 *** Of hard magnetic material

4-dot subgroup: H01F 1/04 **** Metal or alloy

5-dot subgroup: H01F 1/047 ***** Alloy characterized by composition

6-dot subgroup: H01F 1/053 ****** Containing rare-earth metals

The first feature of the patent classification search of the present invention is that when a sub-patent classification code is searched, the top patent classification code of the searched sub-patent classification is provided together as a result thereof. For example, when a "hard magnetic material" is input as a search term, "H01F 1/032. Of a hard magnetic material" is presented as a search result, and H01F 1/032 is referred to as a patent classification symbol mast DB 203), and it is preferable to provide the upper patent classification symbol from the patent classification symbol system tree up to the preset level of H01F 1/032 as a search result. In other words, when H01F 1/032 is input as a search term, it is preferable that the following result is displayed as a search result. It is preferable that the predetermined level is a subclass in case of IPC, a class in case of subclass USPC, a theme in case of FT, or a subclass in case of ECLA or FI, something to do.

Subclass: H01F Magnets

Main group: H01F 1/00 Magnet or magnetic body characterized by magnetic material

1-dot subgroup: H01F 1/01 * Inorganic materials

2-dot subgroup: H01F 1/03 ** characterized by coercivity

3-dot subgroup: H01F 1/032 *** Of hard magnetic material

In order to output the above result, the following steps are performed. First, a keyword entered as a search term is compared with a patent classification symbol index indexed with title information, and the search term is searched for a patent classification symbol as one or more search results included in the title information of the patent classification symbol (When the "hard magnetic material" is inputted, H01F 1/032 is found.) Second, the patent classification code found is referred to the patent classification code mast DB 203, (H01F 1/03, H01F 1/01, H01F 1/00, H01F are found). Third, the above-mentioned top patent classification symbol is found in the patent classification code mast DB 203 And obtains the title information of the found patent classification symbol. Fourth, the patent classification symbol obtained from the first to the third steps and the title information of the patent classification symbol are combined and displayed together with information indicating a layer such as a dot structure.

On the other hand, when a search query is made with "hard magnetic material" AND "rare earths " in the patent classification search, no rows in which" magnet "and" rare earth "are simultaneously present in any of the above rows can not provide search results. However, when considering the patent classification system, "H01F 1/053 ****** containing rare earth metal" should be output as a search result. Therefore, in order to solve the above problems, indexing only on a row-by-row basis does not provide a search result reflecting the systematicity of the patent classification symbol. To do this, we modify the patent classification symbol and title information as follows. The core of the modification is to incorporate all the title information of the upper patent classification symbol of each title into each title information. Table 23 below shows the work type.

[Table 23]

IPC symbol Merged title information H Electricity H01 Electricity; Basic electric device; H01F Electricity; Basic electric device; magnet H01F 1/00 Electricity; Basic electric device; magnet; A magnet or a magnetic substance characterized by a magnetic material H01F 1/01 Electricity; Basic electric device; magnet; A magnet or a magnetic material characterized by a magnetic material; Of inorganic material H01F 1/03 Electricity; Basic electric device; magnet; A magnet or a magnetic material characterized by a magnetic material; Made of inorganic materials; What is characterized by coercivity H01F 1/032 Electricity; Basic electric device; magnet; A magnet or a magnetic material characterized by a magnetic material; Made of inorganic materials; Characterized by coercivity; Of hard magnetic material H01F 1/04 Electricity; Basic electric device; magnet; A magnet or a magnetic material characterized by a magnetic material; Made of inorganic materials; Characterized by coercivity; Of a hard magnetic material; Metal or alloy H01F 1/047 Electricity; Basic electric device; magnet; A magnet or a magnetic material characterized by a magnetic material; Made of inorganic materials; Characterized by coercivity; Of a hard magnetic material; Metal or alloy; Alloy characterized by composition H01F 1/053 Electricity; Basic electric device; magnet; A magnet or a magnetic material characterized by a magnetic material; Made of inorganic materials; Characterized by coercivity; Of a hard magnetic material; Metal or alloy; An alloy characterized by a composition; Containing rare earth metals

Meanwhile, the title information is merged up to the highest patent classification symbol, but it can be merged with a predetermined level black. The predetermined level to be merged is preferably a class in case of IPC, a class in case of subclass USPC, a theme in case of FT, and a subclass in case of ECLA or FI.

When the patent classification symbol and the merged title information are indexed on a row-by-row basis, when "hard magnetic material" AND "rare earth" is obtained as a search term, H01F 1/053 can be found as a search result have. On the other hand, when one patent classification symbol comes out as a search result (for example, when H01F 1/053 comes out), all or a predetermined top patent classification symbol of the search result and title information of the patent classification symbol are searched together The results can be provided as described above. That is, when the search term "hard magnetic material" AND "rare earth" is input, the following search data should be included in the search result.

6-dot subgroup: H01F 1/053 ****** Containing rare-earth metals

If "metal AND coercive force" is entered as a search term, it is preferable that the search result is as follows.

Subclass: H01F Magnets

Main group: H01F 1/00 Magnet or magnetic body characterized by magnetic material

1-dot subgroup: H01F 1/01 * Inorganic materials

2-dot subgroup: H01F 1/03 ** characterized by coercivity

There are means (eg, (+) symbol / mark, etc.) extending around the sub-patent classification of this patent classification around H01F 1/03, and if the user selects this means, It would be desirable to display them hierarchically.

16 is an exemplary diagram of an embodiment in which the patent classification code search module 401 of the present invention is operated. The search includes a step of selecting a patent classification symbol, a step of selecting a search language to search for a patent classification symbol, a step of inputting at least one search word, and a step of performing a search using an inputted search expression to generate a search result .

Meanwhile, the patent classification symbol search module 401 may use the child-patent information between each patent classification symbol to achieve the object of the present invention without utilizing the merged title information. When two or more patent classification symbols appear as a result of query / query entry, they can achieve the same purpose by examining whether there is a child-parent relationship in the patent classification system.

Patent document collection module (404)

Next, the patent document set acquisition module 404 of the present invention will be described. The document collection set access module 404 includes an automatically selected document collection access module 404-1 and a user generated document collection access module 404-2. A document retrieval module 404 through a selection on a directory in which a document set can be obtained through a selection on a directory in which a document set can be specified such as a document retrieval module 404-2-1 through a retrieval expression and an IPC directory 404-2-2). Meanwhile, the patent document set acquisition module 404 may obtain or select a patent document set generated by itself or others, and the patent document set acquisition module 404 may perform a set operation (combination of two or more patent document sets) , A set of differences, a NOT operation, and the like) to generate a new set of patent documents. The document retrieval module 404-2-1 through the retrieval expression is provided with a document collection module 404-2-1-1 through a search engine that queries a search engine and a DBMS 201 The document collection module 404-2-1-2 via the DBMS 201 for querying the document collection module 404-2-1-2. The automatically selected document set acquisition module 404-1 can be automatically obtained from a set of various confirmed patent documents introduced in the determination of the set of patent documents described in the specification of the present invention, except for those generated by the user.

The multi-dimensional analysis operation result table generation module 402

Next, the multi-dimensional analysis operation result table generation module 402 of the present invention will be described. If the data on patent information is constructed in DB, usually the SQL query statement is used to output the desired result. At this time, even if the DB structure is well designed, it is necessary to join a plurality of tables in order to derive a single result, and when each table is large, it takes considerable time to process a select statement.

For example, if the share of the IPC 1 dot subgroup per year is calculated by Samsung Electronics Co., Ltd., and it is composed of a simple select statement, it is possible to use various types of tables (which may differ depending on the design of the DB schema, (Eg, applicant table, IPC table, document table, etc.) should be joined. First, multiple application IPCs should be extracted from SEC's IPC 1-dot subgroup level. Next, we extract the number of submissions per year based on all the patent application data of Samsung Electronics of Korea for the 1 dot subgroup IPC extracted, and then apply the 1 dot subgroup IPCs Find the ratio between the number of applications filed by Samsung Electronics and the total number of applications for each 1 dot subgroup IPC filed 1 dot subgroup of the original 1 dot subgroup Samsung Electronics' share of each IPC is calculated by using one SQL query statement, the length of the SQL query is long, and it takes a long time to process it. On the other hand, in the case of processing the same contents for the LG Electronics, or for the IPC subgroups, it is necessary to perform a large number of calculations in order to obtain the above- There is a possibility that repeatedly varying requirements may be encountered, such as handling the share of the LG Electronics at the level.

In such a case, the patent information is processed in advance so as to meet the various modified requirements, and an operation for multi-dimensional analysis corresponding to the analysis purpose is performed. If table data storing the result is stored, And the response speed is remarkably improved. The multidimensional operation includes a roll-up operation and a cube operation, and the results of such operation performed on the patent data are stored in a DB table. At this time, the table may be referred to as various names such as a cube, a materialized view, and a view, but the contents referred to by the name are the same regardless of the name. The target data is extracted using the SQL query for the above table. In general, multidimensional expression (MDX) queries are used to extract desired data from a multidimensional result table subjected to a cube operation, but they are collectively referred to as SQL for convenience of description. In other words, it is common to query a SQL query to access a table as a result of performing a roll-up operation for multi-dimensional analysis. Accessing a table resulting from performing a cube operation for multidimensional analysis is generally performed using an MDX query, In this specification, both of them have the same nature of accessing the operation result table for multidimensional analysis, and also SQL and MDX query in SQL for convenience of description because they are very similar in the query format. That is, the meaning of querying the operation result table of the multi-dimensional analysis operation result table is as follows: 1) If the table is a roll-up operation, the query is performed by an SQL query; 2) If the table is a cube operation, It should be understood that the query is a query. In the meantime, the term &quot; dimension &quot; means that the analysis is performed on one or more dimensions. (The analysis also includes analysis of one dimension naturally.) In the operation result table for multidimensional analysis of the present invention, When the desired analysis result is logically generated through the n-step information processing, the result of the arithmetic operation from 1 to n-1 or n is stored. Since a conventional SQL query or DBMS has an arithmetic function, it is possible to generate only the result of the intermediate operation in the operation result table for the multi-dimensional analysis, and the final analysis result can be generated through the DBMS or SQL query will be. Hereinafter,

The multidimensional analysis operation result table generation module 402 of the present invention performs a roll-up operation, a cube operation, or other desired operation or data processing on the patent information in order to quickly output desired information, It is an engine that is created as a table. Particularly, when the roll-up, drill down, drill through, or the like occurs, the analysis module of the present invention using the above-mentioned table is configured to generate the multi-dimensional analysis operation result table generation module 402 ) To generate the desired data quickly by performing a simple SQL query on the generated table.

The multidimensional analysis calculation execution result table generation module 402 performs a rollup process for each of the one or more analysis subject categories with respect to the patent document mast DB 202, the patent classification symbol mast DB 203, and the subject mast DB 204, Performs an operation and / or a cube operation, and creates a table as a result of the operation. As described above, the patent document mast DB 202 includes a bibliographic mast DB, and the patent classification symbol mast DB 203 includes a patent classification symbol DB for each type of patent classification symbol system. 204 includes an applicant's name DB, an inventor's name DB and the like, and the applicant's name DB is preferably represented by each country.

The multidimensional analysis calculation execution result table generation module 402 generates the multidimensional analysis calculation execution result table generation module 402 corresponding to at least one analysis subject with data included in the patent document mast DB 202, the patent classification symbol mast DB 203, the subject mast DB 204, In order to generate an analysis result suitable for the analysis subject, a multi-dimensional analysis calculation is performed and the result is generated as a table. There may be a plurality of tables, and the plurality of tables constitute an analysis DW (data warehouse). Preferably, the DB schema of the patent information table used as the material by the multi-dimensional analysis operation result table generation module 402 is constructed with a star schema.

The E-R diagram (Entity Relationship Diagram) of the tables reconstructed with the star schema includes a FACT table in the center and a dimension table as a reference information table that refers to the FACT table. The FACT table includes bibliographic data on the patent document (data in the dimension table is excluded, but there is an ID of each dimension of the dimension table, for example IPC_ID indicating the specific IPC (s) included in the patent document) (At least one date ID corresponding to the date, such as the applicant ID indicating the date (s)) is included in each field. The dimension table includes a table of patent classification symbols (IPC is required, a patent classification symbol for each country in the case of a country-specific FACT table), a date table such as a year, a subject table such as an applicant / agent / inventor, a country / There is a table of other objects that can belong to a dimension, such as the same location table. It should be appreciated that the dimension table includes IDs corresponding to the various IDs included in the FACT table. For example, the IPC dimension table has an IPC_ID, and the IPC_ID is associated with one specific IPC. The tables reconstructed with the above star schema can be basically constructed by country or integrated with countries. It is desirable that the countries that are related to each other, such as family information data, INPADOC / DocDB data, etc., should be managed in a separate table.

The multidimensional analysis operation result table generation module 402 of the present invention generates a multidimensional analysis operation result table for various analysis purposes of the present invention using the data of the patent information table reconstructed as a star schema as a material. Meanwhile, the multidimensional analysis calculation execution result table generation module 402 performs a multidimensional analysis calculation result table creation process using the generated multidimensional analysis calculation execution result table as a material for analysis on another analysis subject or a complex analysis subject You can also create a table. That is, the table of the material for the multidimensional analysis calculation execution result table generation module 402 of the present invention includes 1) a patent information table reconstructed as a star schema, 2) a patent document master database 202, A DB 203, a subject mast DB 204, and 3) a multidimensional analysis calculation execution result table. On the other hand, for convenience of explanation, the above-mentioned 1) to 3) are described as a reconstructed patent information table with the star schema, but it is needless to say that 2) and 3) are not excluded.

The relationship between the operation-result table generation module 402 and the analysis module for multi-dimensional analysis

The multidimensional analysis operation result table generation module 402 of the present invention includes an analysis DW generation module 402-1 for generating the analysis DW, various modules for generating a multidimensional analysis operation result for each analysis subject, There is a module that performs the functions necessary to generate the result of the operation for multidimensional analysis. Various modules for generating a multidimensional analysis operation result for each of the analysis subjects include a multidimensional analysis operation result table generation module 402-2 for total amount analysis, a multidimensional analysis operation result table generation module 402-3 for citation analysis, An operation result table generation module 402-4 for multidimensional analysis for competition analysis, an operation result table generation module 402-5 for multidimensional analysis for inventor analysis, an operation result table for multidimensional analysis for patent technology classification, Module 402-6, a multi-dimensional analysis operation result table generation module 402-7 for fusion analysis, and a multi-dimensional analysis operation result table generation module 402-8 for representative phrase analysis. Each of the modules includes an analysis DW (205-1) for the total amount analysis, a calculation result table 205-2 for the multidimensional analysis for the total amount analysis, a calculation result table 205-3 for the multidimensional analysis for citation analysis, a multidimensional analysis (205-4), inventor analysis multidimensional analysis operation result table (205-5), patent technology classification analysis multidimensional analysis operation result table (205-6), multidimensional analysis for fusion analysis A calculation execution result table 205 -7 for the representative phrase analysis, and an operation execution result table 205 -8 for the multidimensional analysis for the representative phrase analysis.

Next, the relationship between the multi-dimensional analysis operation result table generation module 402 and the analysis module of the present invention will be described. The analysis module includes an analytic expression (SQL query expression) corresponding to at least one analytic subject. The analysis module inquires the operation result table for the multi-dimensional analysis using the SQL query expression, And processes it according to the interface provided by the system 1 of the present invention and provides it to the user of the present system.

Meanwhile, the multidimensional analysis operation execution result table generation module 402 may perform a multidimensional analysis operation execution result table generation process in consideration of the performance of the analysis module (time reduction for obtaining / calculating / acquiring desired data, At least one table may be created for each analysis purpose. That is, the multidimensional analysis calculation execution result table generation module 402 may generate the final analysis result screen data to be provided on the screen of the user, but it is possible to generate only the predetermined level for each analysis subject It is more preferable to generate the final analysis result screen data by using various commands provided by the SQL query itself. This is because, if the final analysis results in a large number of kinds of screens, the latter causes a great inefficiency (waste of computational resources), but if the former method gives up optimization of computational resources (if it is to maximize the response speed) It may be allowed.

Rollups and drilldowns

The concepts of roll up and drill down are described before processing patent information data into the physical view data. Rollups and drilldowns are fundamental to data operations, and are described as examples for ease of understanding. Assume that there is application data for 2006 as shown in Table 24 below.

[Table 24]

2006 January February In March April In May June In July August September October November December Monthly number of applications One 2 3 4 5 6 7 8 9 10 One 2

The above data can be combined by quarter and year as follows. Table 25 below is for conceptual explanations, and the data structure for this roll-up may differ from actual development examples.

[Table 25]

2006 January February In March April In May June In July August September October November December Monthly number of applications One 2 3 4 5 6 7 8 9 10 One 2 quarter First quarter Q2 Third quarter Quarter Quarterly applications 6 15 24 13 year 2006 Number of applications by year 58

In this way, combining small units to larger units in one direction only at one time is called a roll-up operation. That is, an example of a roll-up operation is to combine the monthly unit value with the branch unit value and the branch unit value with the unit value per year, and to access / obtain / extract the result combined into a higher unit in one dimension Called roll-up.

On the other hand, we can see that 58 data in 2006 is composed of 6 data in 1Q, 15 data in 2Q, 24 data in 3Q and 13 data in 4Q. 15 branches, 24 branches in the third quarter, and 13 branches in the fourth quarter (because the original divisions are combined), and the number of branches in each quarter can be divided by month. This approach, called subdivision of a larger unit to a smaller unit in one dimension, to obtain / extract / is called drilldown. Thus, rollups and drilldowns are like two sides of a coin, and if you have multidimensional computed data that performs multidimensional operations such as rollup operations from the smallest unit to the largest unit, you can drill down So that data can be viewed.

Typically, OLAP (online analytic process) is used to access a cube-operated multidimensional analysis operation result table to obtain desired data. A typical OLAP engine supports drill-down and drill-through functions.

Rollups and drilldowns

The following table shows the importance of securing information on subclasses of a patent classification symbol given on IPC and other patent classification schemes. In particular, the present invention, which emphasizes the necessity of obtaining the information of the sub-classifications of the patent classification symbols (for example, patent classification codes of 1 sub subgroups in case of IPC) It becomes possible to realize the importance of the information processing method called automatic inclusion of the sub-patent classification symbol.

Roll-up and drill-down will be described with reference to Table 26 below. Table 33 below shows the number of patent applications of a particular company A centering on H04B 7/02. Parentheses and numbers in parentheses are the number of documents of a specific IPC from patent applications filed by the A company. On the other hand, brackets, numbers in {} are multidimensional numbers. Title information is added for reference, and it is not related to roll-up. However, since it is difficult for users to know what the classification symbol means when they drill down, it is appropriate to provide title information, and multi- (Why the subpopulation classification should be included). The 4-dot subgroup is not listed under H04B 7/02, but is added as a reference to show the multistage nature of the patent classification symbol. Except for the table below, the 4-dot subgroup and title information will be described as a missing table.

[Table 26]

Main group 1-dot subgroup 2-dot subgroup 3-dot subgroup 4-dot subgroup Title / IIPC Description / Patent Classification Symbol Description H04B 7/00 (2000) {2605} Wireless transmission system H04B 7/005 (64) {64} Control of transmission; Equalization H04B 7/01 (4) {4} Reduction of phase shift H04B 7/015 (3) {3} Reduction of echo effect H04B 7/02 (88) {207} . Diversity system H04B 7/04 (57) {114} Using multiple independent antennae spaced apart H04B 7/06 (36) {36} ... in the sending station H04B 7/08 (21) {21} ... in the receiving station H04B 7/10 (3) {3} Using a single aerial system characterized by polarization or directional characteristics H04B 7/12 (2) {2} .. Frequency Diversity System

See H04B 7/02 to understand the rollup. The multidimensional computed number of H04B 7/02 is 207 (= 88 + 114 + 3 + 2), of which 88 is the value of the document containing H04B 7/02 in the patent documents, 114 is the H04B 7/04 and And 3 and 2 are due to H04B 7/06 and H04B 7/08, respectively. On the other hand, the multidimensional computed number 114 (= 57 + 36 + 21) of H04B 7/04, 57 of which is the number of documents containing H04B 7/04 in the patent documents, and 36 and 21 are H04B 7 / 06 and H04B 7/08. On the other hand, the number of documents that are computed multidimensionally in H04B7 / 00 is 2650, which is the numerical value of the documents containing H04B7 / 00 in the patent documents and the multi- ... means that there are more than those shown in Table 33 above among the sub-patent classification symbols of H04B7 / 00.

When retrieving patent document information for H04B7 / 02 (search, counting, or any other statistical, analytical, calculation, etc.), it should be noted that it should not contain only the information marked H04B7 / 02. Because patent information on H04B7 / 02 is more appropriate, it should include patent information on all patent classifications below H04B7 / 02 in the hierarchy of the patent classification system (for example, H04B7 / 02 If you request all documents belonging to the relevant technical domain, it is appropriate to include the documents corresponding to all patent classification under H04B7 / 02, as well as the documents marked H04B7 / 02. ) Therefore, the sub-patent classification of the present invention is the reason why information processing (search, or counting or other statistics, analysis, calculation, etc.) is required. As can be seen in Table 33 above, if there is H04B 7/08 in the IPC information of the patent document, this document also applies to H04B 7/04 and H04B 7/02. Therefore, from the point of view of counting applications or registrations, the counting value for H04B 7/04 is the sum of the values marked H04B 7/04 itself, the values H04B 7/06 and H04B 7/08 Lt; / RTI &gt;

It will be appreciated that the invention may be applied to all levels of roll-up and drill down all IPCs and all other patent classifications as described above. An example of drill down by each IPC would be the following format:

The following tables show one embodiment of drill down by year. If you drill down to the information in Table 33 directly above the year as a year, you can see the data shown in Table 27 below. For the sake of convenience, only the multidimensional computed numbers are left, and the 4-dot subgroup without data is also not displayed.

[Table 27]

Main group 1-dot subgroup 2-dot subgroup 3-dot subgroup H04B 7/00 (200) {2605}

If we drill down to H04B 7/00 with the subpath classification as the reference direction, the data shown in Table 28 below can be seen.

[Table 28]

Main group 1-dot subgroup 2-dot subgroup 3-dot subgroup H04B 7/00 (200) {2605} H04B 7/005 (64) {64} H04B 7/01 (4) {4} H04B 7/015 (3) {3} H04B 7/02 (88) {207}

At this time, if you drill down on H04B 7/015, H04B 7/015 will have no sub-patent classification information, so there will be no further changes. At this time, if you drill down on H04B 7/02, you can see the data as shown in Table 29 below.

[Table 29]

Main group 1-dot subgroup 2-dot subgroup 3-dot subgroup 4-dot subgroup title H04B 7/00 (7) {37} Wireless transmission system H04B 7/005 (64) {64} Control of transmission; Equalization H04B 7/01 (4) {4} Reduction of phase shift H04B 7/015 (3) {3} Reduction of echo effect H04B 7/02 (88) {207} . Diversity system H04B 7/04 (2) {9} Using multiple independent antennae spaced apart H04B 7/10 (2) {2} Using a single aerial system characterized by polarization or directional characteristics H04B 7/12 (3) {3} .. Frequency Diversity System

If you drill down on H04B 7/04, you can see the data shown in Table 30 below.

[Table 30]

Main group 1-dot subgroup 2-dot subgroup 3-dot subgroup H04B 7/00 (200) {2605} H04B 7/005 (64) {64} H04B 7/01 (4) {4} H04B 7/015 (3) {3} H04B 7/02 (88) {207} H04B 7/04 (57) {114} H04B 7/06 (36) {36} H04B 7/08 (21) {21} H04B 7/10 (3) {3} H04B 7/12 (2) {2}

Tables 27 to 37 are the results of roll-up of the filing date / filing year for the specific patent classification symbol of Company A. In other words, the results of each multidimensional computation are from year 1 to year 6 from the past to the year 2000, 2001, 2002, 2003, 2004, 2005 and the present day to the current year (abbreviated as " "). In order to do so, the numerical value of the document corresponding to each patent classification symbol must be rolled up in advance for each set of patent documents of company A, and the following table 31 shows the example.

For the sake of clarity, we describe yearly rollups and drilldowns around H04B 7/02.

[Table 31]

Main group 1-dot subgroup 2-dot subgroup 3-dot subgroup Before 2000 2001 (+) 2002 (+) 2003 (+) 2004 (+) 2005 (+) Recent (2006-present) (+) Total (+) H04B 7/02 (88) {207} 25 15 7 30 60 64 6 207 H04B 7/04 (57) {114} 10 6 3 10 42 39 4 114 H04B 7/06 (36) {36} 3 0 0 3 17 12 One 36 H04B 7/08 (21) {21} 5 3 One 5 4 3 0 21 H04B 7/10 (3) {3} 3 0 0 0 0 0 0 3 H04B 7/12 (2) {2} 0 0 One 0 0 One 0 2

For example, when the user performs a click or the like on the (2001) next to the 2001 (meaning a symbol used as an indication that drilling down is possible for the sake of explanation), the present system 1 (Eg, quarter) data for 2001 and provide it to the user, as shown in Table 32 below.

[Table 32]

Inroup 1-dot subgroup 2-dot subgroup 3-dot subgroup Before 2000 2001 (+) 01 / 1Q 01 / 2Q 01 / 3Q 01 / 4Q 2002 (+) Total (+) H04B 7/02 (88) {207} 25 15 One 2 4 8 7 207 H04B 7/04 (57) {114} 10 6 0 0 6 0 3 114 H04B 7/06 (36) {36} 3 0 0 0 0 0 0 36 H04B 7/08 (21) {21} 5 3 3 0 0 0 One 21 H04B 7/10 (3) {3} 3 0 0 0 0 0 0 3 H04B 7/12 (2) {2} 0 0 0 0 0 0 One 2

Base Direction and Dimension of Rollup

Next, the dimensions used in the present invention will be described in more detail. Typically, the time may be a dimension that is the directional axis of roll-up or drill-down in various units such as one-week-week-month-quarter-year. On the other hand, the patent classification symbol on the patent classification system can also be a dimension that serves as a directional axis of roll-up or drill-down.

First, it is the most basic time dimension. It can be any one or more selected from one-week-week-month-quarter-year-multiple-year unit period and all types of time attributes included in patent documents such as applicant, Field can be applied to the field.

Next, there may be a dimension of the patent classification symbol. At least one patent classification symbol system is introduced for each country, and the dimension of the patent classification symbol can be set to reflect the multistage property of the patent classification symbol system itself . Meanwhile, a bundle of the at least one selected patent classification symbol may be set as a unit, and another dimension may be set. For example, a multi-level subchapier corresponding to the subject can be set in a multistage manner on a large topic of RFID, and a separate dimension can be set by a method of associating the bundles of the patent classification symbols with the subchapters. These separate dimensions are particularly useful for personalized multidimensional analytical results-based performance tables / personalized cubes - tables / cubes that each individual user makes to achieve his or her analytical purpose.

The region can also be a dimension. Among the regional units, the most easily obtainable unit from the patent document is the country, and from the address information, it is possible to create a dimension having a multi-level hierarchy by dividing into regions within one country.

The subject can also be a dimension. The subject includes applicants, inventors, and agents, and quantitative attributes such as applicant's attributes (enterprise, research institute, university, etc.), size (large, medium, small, etc.), and attributes between applicants such as group- . On the other hand, corporate financial information such as stock prices, sales, and profit rate of companies and corporate financial information can be a dimension. On the other hand, meta-attributes such as global companies and local companies, which can be arbitrarily assigned to each company, can also be a dimension.

On the other hand, attributes can also be a dimension to the status of documents such as application, registration, rejection, etc. On the other hand, when grouping various counting values, the group to which each counting value belongs may also be a dimension. Examples of such groups include claims 1 to 5, 6 to 10, 10 to 15, 15 or more, and the number of co-applicants, the number of co-inventors, the number of families (the number of domestic families, Number, etc.) The number itself can not be a dimension, but when the number is grouped, the group to which it belongs can be a dimension.

The multidimensional analysis operation result table generation module 402 of the present invention generates a multidimensional analysis operation result table by performing a multidimensional operation for each selected at least one dimension. When the multidimensional analysis calculation execution result table generation module 402 of the present invention generates the multidimensional analysis calculation execution result table, if the patent classification symbol is included in the dimension, the multidimensional calculation operation for the given patent classification symbol Value is generated, the multilevel calculation value is generated by considering the patent classification symbol and the upper patent classification symbol of the patent classification symbol together. When the multidimensional operation value is generated, when a multidimensional operation value for an arbitrary patent classification symbol is obtained, all values of the patent classification symbol and its subpublic classification symbols are multiplied by the multidimensional computed numerical values, . The multidimensional analysis operation result table generation module 402 of the present invention generates the multidimensional analysis operation result table generation module 402 in accordance with the patent classification notation And reflects a value obtained by performing a multidimensional operation on the symbol. For example, when document number # 1 is assigned IPC H04B 7/06, when the multidimensional operation data is generated by this document, H04B 7/06 is also given a counting value of 1, and H04B 7/06 Count values must also be assigned to H04B 7/04 and H04B 7/02 immediately above. Of course, it is common sense that counting values are also given to higher levels of H04B 7/00 and higher.

As such, dimensions that are the axis of roll-up and drill-down are mainly dimensions. Patent categorization symbols such as IPC, time, etc. are typical examples of such dimensions, and include countries, regions, applicants, inventors, Information can also be a dimension axis. The multidimensional analysis calculation execution result table generation module 402 of the present invention calculates the analysis index values such as the total amount, the occupancy rate, the concentration rate, the activity rate, and the like, using the selected at least one dimension as the target axis of the rollup.

In the first place, the whole country - the individual country - the area within the individual country (the patent information includes the country and / or address information of the applicant and / or the inventor) ..., and so on. This will be the same way - time-period-period.

If the applicant is subdivided, it will be possible to subdivide it into applicant (national integration) - applicant for each country, and it will be possible to integrate all national integration and intermediate regions (North America, Asia, Europe, Middle East, South America, Africa, etc.) . In order to introduce medium scale, it is necessary to map a country name corresponding to North America (for example, USA, Canada, Mexico, etc.) to an intermediate category called North America.

In addition, the applicant to the inventor can be rolled up / drilled down. The direction may be as follows. 1) Applicant (National Integration) - Applicant - Rollup / Drill down to each country's inventor or 2) Applicant (National Integration) - Inventor (National Integration) On the other hand, it will be possible to roll-up / drill down according to the attribute of the applicant in one country. For example, applicants are divided into organizational type units such as corporations, universities, and research institutes. The companies are divided into large corporations, mid-sized corporations, and small and medium-sized enterprises, and universities are divided into national and public universities, private universities, And then roll-up / drill down on them. At this time, it is natural that there should be mapping information for each applicant. For example, information must be mapped that applicant A is a corporation and its size belongs to a small business.

At this time, if there is financial statement data for a company, more variety of roll-up / drill-down can be done. For example, if there is information on sales, profit, share price, and each increase / decrease ratio in company financial statements, each company - sales scale - revenue scale, company - profit - profit increase rate, or company - stock price - You can roll-down / drill down into various categories such as scope. That is, the essence of roll-up / drill-down is to arrange the information processing results for each category when there is a plurality of category information for one object. In one patent document, there is country, time, applicant, inventor, patent classification information, and applicant information has local information of the applicant. If the applicant is a corporation, it may be constructed as an enterprise information DB in which the company's financial statement information may be included. At this time, if a selected category has a multi-level hierarchical structure such as 1) patent classification symbol information in whole or set arbitrary set of patent documents, various information processing (search, or the like) corresponding to the patent classification symbol selected in the category Counting, or any other statistic, analysis, calculation, etc.), and output the result. 2) Various intersections of two or more categories. (When the category type is n, the number of possible intersections is nCr, (n combination r, r is at least 1 but less than or equal to n), but not all of these combinations may be selectable, and it is natural that they have special analytical meaning and combinations. Roll-down / drill-down on the basis of &quot;

The following points should be noted when rolling / drilling down through an intersection.

The first is when you use two categories. For example, if you rolled up / drill down to category A (for example time) and category B (for example IPC), you can see the results of information processing (search or counting or other statistics, analysis, For convenience of presentation, it is desirable to drill down in one direction (for example, in the detail IPC, the time axis is not drilled down, only the IPC is unfolded (this is exemplified when explaining the roll-up / drill down concept above) , IPC can also be drilled down and the time axis can be drilled down, but in this case the result may seem too complex, so it is desirable to avoid it unless it is a special case. , The roll-up can be performed in the A category direction of each cell based on the information value of the cell composed of all the combinations of the predetermined unit of category A and the set unit of category B The table immediately above shows an example. Each value in the cell corresponds to the information value of the cell, and the { Corresponds to rolling up the number of patents per year from 2000 to 2005 at each patent classification level, 17 in H04B 7/02 {17} corresponds to H04B 7/02 and its sub-patent classification The number of patent documents up to the symbol is rolled up. In the table above, roll-up based on the year (for example, how many patent documents are in 2005, how many are in H04B 7/02 in 2005, etc. , But it would also be natural that it could also be rolled up and displayed, and it would of course be multidimensional computed on a quarterly and monthly basis).

The second is when using more than three categories. In this case, it is preferable that the drill-down is expanded in one direction to be selected, and it is acceptable to expand in two directions. However, it is preferable to expand in three directions because it is impossible to express visually or extremely difficult. In this case, it is preferable that the roll-up is calculated for all categories, but it is also preferable to process the dimension axis that does not need to be rolled up beforehand according to the roll-up calculation amount and frequency of use, . The necessary information can be rolled up and processed at the time of occurrence even without roll-up, but a relatively long time may be consumed for roll-up processing. For example, if you do not roll-up by year, you can obtain year information from bibliographies of all the documents you are interested in, calculate separately for each year, and print the year by year, . However, it is advantageous in terms of response speed when it is calculated by dividing it in advance by the year and rolling up the roll-up by the year.

Total amount analysis Multidimensional analysis Operation result table data

Total amount analysis Multidimensional analysis Operation result table data

Hereinafter, the multi-dimensional analysis operation result table generated by the multi-dimensional analysis operation result table generation module 402 of the present invention will be described in more detail. In addition, how the analysis module accesses the multidimensional analysis operation result table and what data is generated will also be described.

In general, since the date of the filing date indicated in the patent document is denoted by the date of the month, if the number is counted by each patent classification symbol, the multi-dimensional operation can be performed on a month unit, a branch unit or a podium.

Table 33 below shows a portion of one embodiment of a result table in which a multidimensional operation is performed. The data as shown in the following Table 33 is generated by the multidimensional analysis calculation execution result table generation module 402 in the patent information table reconstructed by the star schema based on the date of the IPC and the year of the applicant, It is the data created by counting the number of patent documents for each layer of multi-level IPC. Data such as Table 33 can be stored in any format such as table / view / materialized view, and the data contained in the stored table can be extracted using the appropriate SQL query statement. For example, the number of applications for each IPC (multidimensional computed year) for each IPC 1-dot subgroup reference (C5 level in the following) of Applicant A is taken as an example. (It is better to have the grand total data for the year for this purpose.) (In Table 33, the space is not displayed due to space constraints, The generation module 402 refers to the patent classification symbol DB of the IPC and the like and selects the patent classification code corresponding to the sub-patent classification code of the patent classification code that can not be accessed by using a wild card (*,? (This is especially important at the level of the dot in the title information, even for sub-groups for IPC, and for FT below class level for USPC)

[Table 33]

GC C5 C6 C7 C .. 01 02 03 04 05 H04B 7/02 H04B 7/04 H04B 7/06 0 0 3 17 12 H04B 7/02 H04B 7/04 H04B 7/06 0 0 3 17 12 H04B 7/02 H04B 7/04 H04B 7/08 3 One 5 4 3 H04B 7/02 H04B 7/04 H04B 7/08 3 One 5 4 3 H04B 7/02 H04B 7/04 3 2 2 21 24 H04B 7/02 H04B 7/04 3 2 2 21 24 H04B 7/02 H04B 7/04 6 3 10 42 39 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 9 3 20 18 24 H04B 7/02 9 3 20 18 24 H04B 7/02 9 3 20 18 24 H04B 7/02 15 7 30 60 64

The table will be described as follows. AppName is the name of the application, the number under C is the node layer from the root node of IPC (C1 is the section, C2 is the class, C3 is the subclass, C4 is the main group, C5 is the 1 dot subgroup, 2 dot subgroup, and C7 is a 3-dot subgroup, which may be assigned to C8 through C20, usually up to C15), and 01, 02, and so on. The year is before 00, and the numbers are omitted. If the present data is 07, the numerical value based on the filing date based on the documents published so far is the data value of each cell.

On the other hand, when generating a value in the table, the value of each cell of the sum field may be calculated by summing up the value of each year, and the gray period may be one year and six months The period from the previous day to the present day as a unit period and take the sum of the values of each year belonging to the period as the unit period.

The GID is used to show to what level the result of the multidimensional operation. In the case of GID 7 H04B 7/02, the result of multidimensional operation (for example, 64 = 24 + 39 + 0 + 1) (Except oneself) to show the result of multi-dimensional computation. 24 is the value from GID 1, and 24 is the number of the patent document with the patent classification symbol H04B 7/02. The representation of the GID value may be in any manner (arbitrary), but in this specification it is denoted by a sequence of "

GID is the rollup stage. Roll up to C8 level is rolled up to GID 0, C7 level is rolled up to GID 1, Roll up rolled up to C6 level is rolled up to GID 3, Roll up rolled up to C5 level is rolled up to GID 7, Is rolled up to the level GID 15, and the roll up to the C3 level is denoted as GID 31, and the roll up to the level C 2 is denoted as GID 63. At this time, it is obvious that the GID can be allocated even before the C8 level and at the C1 level, but it is not necessary to reflect the GID to the system 1 because it is not very effective. In H04B 7/06, GID 0 rolls up to C8 level and GID 1 rolls up to C7 level. (In the example above, rolled up to C7 and rolled up to C8 are the same because there is no value in C8.)

The correspondence relationship between the GID and the level of the patent classification symbol used in this specification is described as one example, and the description will be centered on Table 34 below.

[Table 34]

GID Target range IPC level USPC level FT level ECLA Level 0 C8 4 dot 4 dot 4 dot 4 dot One C7 3dot 3dot 3dot 3dot 3 C6 2dot 2dot 2dot 2dot 7 C5 1dot 1dot 1dot 1dot 15 C4 main group 0dot 00 Level main group 31 C3 subclass class Theme code subclass 63 C2 class super category super category class 127 Applicant Applicant Applicant Applicant Applicant 255 all all all all all

The GID is a code that means that it rolls up to its corresponding level. For example, GID 15 is at the C4 level, which means that rollups (roll up to themselves and all subclasses of their own) from the main group to the main group, the USPC to 0dot, the FT to the 00 level, and the ECLA to the main group, As shown in FIG. Therefore, if you want to view information at a certain level, you can specify the GID corresponding to that level, and you can see the multidimensional computed information up to that level. In the present specification, the abbreviation GID is used, but it can be referred to as a "roll-up level code" In the above and below, the rollup level code and the GID are used in the same sense. In the FT, the 00 level usually refers to a level with two alphabetic characters, such as AA, which is not followed by a number. In the case of USPC or FT, the super category means that there are too many items of the class and the theme code level, and these are grouped by category.

As can be seen from Table 33, it can be seen that each patent classification symbol in each row in the table has its own top patent classification symbol on the left side. The multidimensional analysis operation result generation module of the present invention generates multidimensional computed numeric values for each of the IPC symbols as data shown in Table 33. [ This can be created for every patent classification set in each patent document set for a set of all patent documents for a particular applicant A by country. On the other hand, data as shown in Table 33 and Table 33 may be generated based on all application documents, or may be generated based on all registered documents.

The multidimensional analysis operation result table generation module 402 may generate a multidimensional analysis operation result table as shown in Table 33 for each document set having a predetermined property. The multidimensional analysis operation result table generation module 402 generates a multidimensional analysis operation result table 301 by using a key for identifying a patent document such as an application number of a patent document included in a document set having the predetermined property, Extracting only the key values having the key value that can specify the patent document in the FACT table of the table reconstructed with the star schema structure, 3) extracting the patent document corresponding to the extracted key values And generating a multidimensional analysis operation execution result table according to at least one predetermined analysis subject with respect to only the predetermined analysis subject. After the steps 1) and 2) are performed, the multi-dimensional analysis operation result table generation module 402 generates a target document set to be subjected to predetermined processing. The FACT table includes FACT information for the entire set of documents. The FACT information naturally includes document specific information, and may include various bibliographic information. Preferably, the document specific information is an application number or a document specific code or a document serial code.

The multidimensional analysis calculation execution result table generation module 402 generates at least one multidimensional analysis calculation execution result table for each document by country or country integration according to a predetermined analysis theme by at least one predetermined analysis subject . The arbitrary document set may be any set of documents sharing at least one attribute that can be defined in advance. Examples of such arbitrary document set types include: 1) a collection of patent documents generated by a method of specifying a set of patent documents as a specific patent classification symbol on a specific patent classification system in a specific country DB, 2) (3) the name of a specific inventor (ie, applicant's name and inventor's name) included as an inventor in the patent document of a specific applicant in a specific country DB, 4) a set of patent documents generated by a method of specifying a set of patent documents by a specific agent name in a specific country DB, 5) a set of patent documents generated by a method of specifying a set of patent documents, By specifying a specific patent classification on the patent and by specifying a set of patent documents by the specific applicant's name 6) a set of patent documents generated by a method of specifying a set of patent documents by a specific patent classification mark on the specific patent classification system, a specific applicant's name and a specific inventor's name in a specific country DB, 7) 8) a set of patent documents generated by a method of specifying a set of patent documents by a specific applicant's name and a specific agent's name in a particular country DB, 9) a set of all patent documents in a particular country, 9) 10) An example of a set of patent documents specified in the above-mentioned 1) to 9) in units of a predetermined period. On the other hand, the above-mentioned arbitrary document set may be created only for the family information in the foreign countries other than the first country, or in the case of the United States, a document set consisting only of the reissued document. The document set referred to in this paragraph is exemplary, and with respect to these illustrative examples, the multi-dimensional analysis operation result table generation module 402 of the present invention may include at least one It is preferable that at least one or more of the multidimensional analysis operation execution result tables are generated for each set of analysis topics.

The multidimensional analysis operation result table generation module 402 may generate a multidimensional analysis operation result table as shown in Table 33 for any document set generated by the user. The multidimensional analysis operation result table generation module 402 generates a multidimensional analysis result table generation module 402 for generating a multidimensional analysis result table based on a key for identifying a patent document such as an application number of a patent document included in an arbitrary document set generated by the user, 2) extracting only key values having a key value that can specify a patent document in the FACT table of the table reconstructed with the star schema structure, 3) extracting key values corresponding to the extracted key values And a step of generating an operation result table for multidimensional analysis for at least one or more predetermined analysis topics only for documents.

A processing method of an operation execution result table generation module 402 for multidimensional analysis of the total amount data

In order to generate a multidimensional analysis operation result table as shown in the above table, the multidimensional analysis operation result table generation module 402 generates the multidimensional analysis result table for generating a multidimensional analysis operation result table (e.g., Lt; / RTI &gt; the following steps are performed.

First, a key (key unique information) value that can specify a patent document such as an application number of a patent document included in the document set having the predetermined property is obtained, and a table reconstructed with the star schema structure A key table having a key value for specifying a patent document is extracted from the FACT table of the key table, and at least one set of instructions for a basic table for a multidimensional calculation by a predetermined analysis subject (only the patent documents corresponding to the extracted key values (The command set may be a script.) (At this time, a multidimensional operation execution instruction set for each analysis subject for the multidimensional operation base table may also be obtained at the same time.) If an application or registration The entire set of patent documents consisting of all the documents, Obtain only a set of instructional instructions for the analysis subject to generate the operation result table for the multi-dimensional analysis according to the predetermined analysis topic. That is, in this step, the multidimensional analysis operation execution result table generation module 402 determines a processing target document set for performing the predetermined processing, and the instruction set for each analysis topic for the predetermined processing is determined.

Second, a base table for the multidimensional operation is generated by executing an instruction set according to an analysis topic for each document of the target document set. The base table for the multi-dimensional operation includes information on at least one dimension, and basic data for each dimension for performing a multi-dimensional operation is input to each of the records. The basic data includes information on the presence or absence (1 or null) of information (application / registration has occurred, was absent) and various counting information (any one or more of the information referred to herein as counting information, May be included. The dimensions include: 1) a patent classification symbol for each type of patent classification symbol, 2) a date dimension, 3) a location dimension (country, region, etc.), 4) an attribute dimension of a subject It is possible to divide the applicant into companies, universities, research institutes, individuals and others, and divide the enterprise into multinational corporations, large corporations, small and medium enterprises, (Refer to the attributes of each applicant in the above), or a composite dimension resulting from combination of any of the above.

At this time, a field must be designed in the base table for the multidimensional operation in order to input a bibliography necessary for each dimension. The field contains a patent classification symbol dimension that can accommodate all upper and lower patent classification symbols in the patent classification scheme, such as from IPT to n dot subgroups, from USPC (from super category) to n dot levels There should be a field for each level. On the other hand, for the time dimension, it is sufficient to have the smallest unit field for analysis. In the case of a patent analysis, it would suffice for a month field only if a day or week analysis is not required. On the other hand, if there is an applicant attribute dimension, there should be fields that correspond to companies (multinational corporations, large corporations, small and medium enterprises), universities, researchers, individuals,

For example, if it is sufficient that only the patent classification symbol dimension and the filing date dimension based on the applicant is sufficient in the basis table for the above-mentioned multi-dimensional operation, the applicant A filed on January 3, 2005, and the IPC has applied IPC to H04B 7/06 The data as shown in Table 35 below should be generated for the document.

[Table 35]

Station C1 C2 C3 C4 C5 C6 C7 05/01 05/02 05/03 H H04 H04B H04B 7/00 H04B 7/02 H04B 7/04 H04B 7/06 One

On the other hand, the period and applicant attributes may be entered as follows. The data as shown in Table 36 below is a partial multidimensional operation performed in the year and the applicant attributes, and may be processed in the multidimensional operation described later, but it is not a problem even if it is created in advance.

[Table 36]

C1 C2 C3 C4 C5 C6 C7 05 05/01 05/02 05/03 H H04 H04B H04B 7/00 H04B 7/02 H04B 7/04 H04B 7/06 One One

On the other hand, there may be further included corporate evaluations such as the applicant's position or applicant's financial level, the number of families based on the number of families, and the number of claims based on the number of claims. As well as additional data.

[Table 37]

country State City Number of families Category 1 Number of families Category 2 Company Ratings n-1 Company Ratings n-2 Claim range 1 Claim range 2 Kr Gyeonggi-do Suwon One One One

A method of generating the above data is as follows. 1) The patent classification symbol included in the document is referred to the patent classification code mast DB 203 to obtain all the upper patent classification codes of the patent classification codes, . For example, when IPC is H04B 7/06, H04B 7/06 is assigned to the 3-dot subgroup (C7) level, H04B 7/04 is assigned to the 2-dot subgroup (C6) / 02, H04B 7/00 in the main group C4 level, H04B in the subclass C3 level, H04 in the class C2 level, and H in the section C1 level. 2) Enter the required value in the time dimension based on the required reference date (here, filing date) from various date information such as filing date / registrant. If the filing date is the base date, January 3, 2005, and the time dimension is divided into dimensions of month, quarter, and year, enter 1 in January of 2005. 3) Enter other bibliographies in the other dimension. Then, enter 1 into a large corporation. The field value such as the applicant's position is filled using the address information, the patent document master DB 202 is inquired to inquire the number of the family of the patent document to be a subject, the category to which the inquiry value falls, &Lt; / RTI &gt; The claims range is treated in the same way as the number of families.

When there are two or more kinds of patent classification symbols in one document (for example, IPC and USPC), a base table for the above-mentioned multi-dimensional calculation may be generated for each type, and two types or more of patents You may also want to process the classifier. On the other hand, when a plurality of patent classification codes of the same kind are included in the document, it is desirable to generate an independent record for each patent classification code. On the other hand, if the applicant is two or more co-applicants, a record having the same information for each applicant is generated for the number of applicants (if the subject is an inventor, a record is generated for each inventor, and for agent analysis, Create records for each agent.)

Third, a multidimensional operation is performed on the basis table for the multidimensional operation, and a multidimensional analysis operation execution result table is generated. The data of the table as shown in Table 38 below shows the result of performing the roll-up operation from the IPC dimension to the time dimension up to the year dimension. (For the sake of convenience, it is shown that only the year 2004 is rolled up to the quarter- .)

[Table 38]

C5 C6 C7 C .. 01 02 03 04 04 / 1Q 04 / 2Q 04 / 3Q 04 / 4Q H04B 7/02 H04B 7/04 H04B 7/06 0 0 3 17 One 2 3 11 H04B 7/02 H04B 7/04 H04B 7/06 0 0 3 17 One 2 3 11 H04B 7/02 H04B 7/04 H04B 7/08 3 One 5 4 One 2 One 0 H04B 7/02 H04B 7/04 H04B 7/08 3 One 5 4 One 2 One 0 H04B 7/02 H04B 7/04 3 2 2 21 4 4 4 9 H04B 7/02 H04B 7/04 3 2 2 21 4 4 4 9 H04B 7/02 H04B 7/04 6 3 10 42 6 8 8 20 H04B 7/02 H04B 7/10 0 0 0 0 H04B 7/02 H04B 7/10 0 0 0 0 H04B 7/02 H04B 7/10 0 0 0 0 H04B 7/02 H04B 7/12 0 One 0 0 H04B 7/02 H04B 7/12 0 One 0 0 H04B 7/02 H04B 7/12 0 One 0 0 H04B 7/02 9 3 20 18 6 6 6 0 H04B 7/02 9 3 20 18 6 6 6 0 H04B 7/02 9 3 20 18 6 6 6 0 H04B 7/02 15 7 30 60 12 14 14 20

The multidimensional operation is divided into a roll-up operation and a cube operation. The roll-up operation and the cube operation are described based on the IPC, the date, and the applicant, respectively. When the IPC, date, and applicant dimensions are present, the rollup operation proceeds only in one of the three dimensions, while the cube operation is 3P3 (the number of permutations generated when three of the three are selected, here 6) Lt; / RTI > That is, when the IPC dimension is 1, the date dimension is 2, and the applicant dimension is 3, the roll-up operation proceeds only in one direction such as 1-> 2-> 3, But not in all six directions, such as 1-> 3-> 2, 2-> 1-> 3, 2-> 3-> 1, 3-> 1-> 2, 3-> 2-> Or more. Therefore, when the roll-up operation is performed, if the result of the operation up to 1-> 2-> 3 is not needed (the final result of the rolled up operation in the applicant's type unit is generated) Alternatively, the roll-up operation may be performed by designating a desired roll-up direction only for at least one of dimensions selected when IPC and date dimensions are rolled up to the date dimension as shown in Table 38. [

The multidimensional analysis calculation execution result table generation module 402 obtains the filing date information and the patent classification symbol information from the given set of patent documents and then obtains the patent classification symbol mast DB 203 or star And extracts the upper patent classification symbol of the obtained patent classification symbol with reference to the table reconstructed by the schema structure. The data of the multidimensional analysis operation execution result table including information such as the above table is extracted from the extracted upper patent classification symbol and year information of the filing date. At this time, if there are two or more kinds of patent classification symbols in one patent document (for example, in the case of the United States, there are IPC and USPC), the classification of each patent classification may be independently performed. If a patent document contains two or more patent classification codes of a single kind, it may be necessary to 1) process only the first main patent classification code, 2) process all patent classification codes, or 3) If weights for classifiers and subpublic classifiers are present, they can be processed to reflect the weights. In this case, in the case of 3), a decimal point may appear in the numerical value, and in this case, rounding is a valid notation. At this time, it is a matter of choice to take any one of the above 1), 2) and 3). It should be noted that other policies other than the above 1), 2) and 3) may be possible. When the method 2) is adopted, the number of lines of the operation result table for multidimensional analysis increases, because at least one new data value is generated for each patent classification symbol (that is, for the sub IPC, It should be done). On the other hand, when the method 3) is adopted, the numerical value can be expressed as a decimal point.

The table is an indicator of the number of patents, such as the number of applications or the number of registrations. The data processing such as roll-up / drill-down including the sub-patent classification symbol of the present invention may be utilized in various patent information analysis indexes such as occupancy rate, concentration rate, activity index (AI)

Perform multidimensional analysis operations without a base table for multidimensional operations by analysis topic.

In the above description, generation of a multidimensional analysis operation execution result table using the basis table for multidimensional operation by the analysis subject has been described. The multidimensional analysis operation result table generation module 402 of the present invention generates the multidimensional analysis result table generation module 402 for the multidimensional analysis operation result You can also create a table. The base table for multi-dimensional arithmetic operations according to the analysis subject functions to reduce the amount of computation / information processing of the arithmetic operation result table generation module 402 for multi-dimensional arithmetic operations. One of the base tables for multi- As shown in FIG. The multidimensional analysis operation execution result table generation module 402 obtains necessary information from a table re-aggregated into the star schema structure even when there is no base table for multidimensional operation by the analysis subject, processes the multidimensional analysis result table, .

Performs multidimensional analysis operations without table reconstructed with star schema structure Generates result table

The multidimensional analysis operation result table generation module 402 generates the multidimensional analysis operation result table for generating the multidimensional analysis operation result table 402 by using the patent document mast DB 202, the patent classification symbol mast DB 203 and / or the subject mast DB 204 ) And generate a basis table for the multidimensional operation by the analysis subject. In this case, the program (script, script, etc.) for generating the basic table for the multi-dimensional calculation according to the analysis subject may be relatively complicated and the recyclability of the entire program or each module constituting the program may be relatively low There will be. For example, in the base table for the multi-dimensional calculation according to the analysis subject, information on the IPC itself and all IPCs of the IPC itself is stored for each application number. The information is stored in the multi- (402) finds the IPC included in the application number, inquires the IPC found in the patent classification code mast DB (203) or other top patent classification code table and the like, The base table for the multi-dimensional calculation according to the analysis subject can be generated based on the obtained information.

Patent document Creation of operation result table for multidimensional analysis by analytic subject from mast DB (202)

In order to generate the multidimensional analysis operation result table, the multidimensional analysis operation result table generation module 402 may generate the multidimensional analysis operation result table by using the predetermined / acquired document set (for example, a set of all documents registered or registered in a specific country) The following steps are performed.

First, a key (document unique information) value for specifying a patent document such as an application number of a patent document included in the document set having the predetermined attribute is obtained and stored. Meanwhile, in order to generate the multidimensional analysis operation result table for the analysis subject as shown in Table 38, at least one data among the data of Tables 35 to 37 as a material is stored in the patent document mast DB (202), the patent classification symbol mast DB (203) and / or the subject mast DB (204), integrates the generated data, and stores it in the memory. The data stored in the memory may be data equivalent to the data stored in the base table for the multi-dimensional operation. That is, the data stored in the memory includes information on at least one dimension, and basic data for each dimension for performing a multidimensional operation is input to each of the records. The basic data includes information on the presence or absence (1 or null) of information (application / registration has occurred, was absent) and various counting information (any one or more of the information referred to herein as counting information, May be included. The dimensions include: 1) a patent classification symbol for each type of patent classification symbol, 2) a date dimension, 3) a location dimension (country, region, etc.), 4) an attribute dimension of a subject It is possible to divide the applicant into companies, universities, research institutes, individuals and others, and divide the enterprise into multinational corporations, large corporations, small and medium enterprises, (Refer to the attributes of each applicant in the above), or a composite dimension resulting from combination of any of the above.

The multidimensional analysis calculation execution result table generation module 402 analyzes the data stored in the memory in which at least one of the information in Tables 35 to 44 for each patent document is integrated, And generates an operation result table for the multi-dimensional analysis according to the analysis subject as shown in Table 38. [

Ranking information by IPC level by applicant

Given the multidimensional analysis operation result table as shown in Table 38, the analysis module can generate the following analysis result.

First, when the applicant is acquired, the applicant can generate ranking information based on the number of applications / registration number per level of the patent classification symbol for the applicant. The ranking information may generate ranking information by a method of comparing count values of the multidimensional calculated application or registered documents for each level of at least one patent classification symbol of the acquired applicant. For example, in the multidimensional analysis operation result table data, the count value of the multi-dimensional calculated document of H04B 7/02 at the 1dot subgroup level (C5 level) of applicant A and the count value of the multidimensional document of the applicant A at the same 1dot subgroup level The multi-application / multi-registration ranking of each patent classification symbol at the 1-dot subgroup level of applicant A can be obtained by comparing the number of multi-dimensional calculated documents corresponding to the other patent classification symbols of patent A .

Total amount analysis module

Assume that there is a multidimensional analysis operation execution result table for each applicant, IPC level, and year as filed (or registration) count as shown in Table 39 below.

[Table 39]

GC C4 C5 C6 C7 C .. 01 02 03 04 05 H04B 7/00 H04B 7/02 H04B 7/04 H04B 7/06 0 0 3 17 12 H04B 7/02 H04B 7/04 H04B 7/06 0 0 3 17 12 H04B 7/02 H04B 7/04 H04B 7/08 3 One 5 4 3 H04B 7/02 H04B 7/04 H04B 7/08 3 One 5 4 3 H04B 7/02 H04B 7/04 3 2 2 21 24 H04B 7/02 H04B 7/04 3 2 2 21 24 H04B 7/02 H04B 7/04 6 3 10 42 39 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 9 3 20 18 24 H04B 7/02 9 3 20 18 24 H04B 7/02 9 3 20 18 24 H04B 7/02 15 7 30 60 64 B29C / 100 .. .. H04B ..

The multidimensional analysis operation result table may be generated based on the applicant for all applicants in a specific country. For example, if there is a patent classification symbol and a filing date (date) included in a specific patent document of the applicant C, the patent classification system and the filing date data of the patent document are referred to, (Such as a patent classification code, an applicant (AppName), etc.), or increase the number. As described above, when generating multidimensional analysis operation execution result tables for all applicants in a specific country, the analysis module can generate the following analysis results from the multidimensional analysis operation execution result table.

To show applicant A's IPC subclass level grouped information in the first place, a yearly count value at this level may be generated. The value of the most recent field is the day of the year in which one year and six months (18 months) of transition belongs to the preset reference date (the date of inquiry of the operation result table for multidimensional analysis) It is preferable that the count value is derived from the patent document obtained up to the inquiry result table inquiry date. (... refers to the corresponding counting value). Table 40 shows an example thereof.

[Table 40]

ranking IPC subclass 01 02 03 04 05 lately One H04B (+) 2 B29C (+)

Of course, if we drill down on H04B of Table 40 above, the count value for the documents corresponding to the patent classification symbol that is above the patent classification symbol of H04B among the patent documents of applicant A will be as shown in Table 41 below. This is done by counting on the condition of "AppName = A and IPC level = C4" in the operation result table for multidimensional analysis as shown in the above table, and ordering will be based on C4 (order by IPC level = C4, C4 Means the IPC subclass.)

[Table 41]

ranking IPC subclass IPC main group 01 02 03 04 05 lately One H04B (+) H04B 1/00 (+) H04B 3/00 (+) B29C (+)

If drilling down again in the above Table 41, the multidimensional computed count value on the 1-dot subgroup basis will be as described immediately before. (Of course, if you drill down on H04B, the count value for the documents that fall under the patent classification mark that is above the patent classification mark H04B among the patent documents of Applicant A will be as follows: (IPC level = C5, C5 is the IPC main group), and the IPC level (C5, C5) will be counted as "AppName = A and IPC level = C5" ). The counting value in each cell is a value obtained by multi-dimensional calculation of all values corresponding to the sub-patent classification code of the cell.

If you drill down on H04B 1/00 (+) above, you will get the patent classification code below it, and the value of each cell will be counted under the condition "AppName = A and IPC level = C6" C6 (order by IPC level = C6, C6 means IPC 1 dot group). The counting value in each cell is calculated by multiplying all the values corresponding to the sub-patent classification code of the cell Value. In this way, it is possible to continue to drill down until there are no more sub-patent classification symbols, and the necessary cell values can be generated by inputting the above conditions until further drill down is not possible.

Although the above method has been described for IPC, it will be obvious that the same applies to USPC, FT, FI, and ECLA.

Similarly, the above describes a method of generating a multi-dimensional arithmetic operation result table for multidimensional analysis, a method of calculating a cell value at the time of drilling down, and a method of expressing a calculated value based on the application (or the total amount of registration) / RTI &gt; and / or the activity rate, as will be apparent to those skilled in the art.

On the other hand, it is possible to generate a multidimensional calculation value in the same manner for another analysis index corresponding to the predetermined defining formula in which the patent classification symbol is intervened in the above manner, and the cell value at the time of drilling down can be calculated , And the calculated value can be expressed.

When there is an operation result table for the multidimensional analysis as described above, when at least one specific patent classification symbol is given, the rank value of the applicants and the applicant's yearly count value may be generated with a high count value. For example, assuming that H04B is given, it will be done by applicant to order by IPC level = C4 and C4 = H04B.

Share and other data generation examples

For example, if the occupancy rate is defined as "a value obtained by dividing the number of documents of a set of objective documents having a specific attribute by the number of documents of a total document set having a specific attribute", data of a table format as shown in Table 39 can be generated will be. That is, if the occupancy rate of the specific patent classification code of the specific company A is obtained, the target document set is a document set including a specific patent classification symbol among A document sets, and the entire document set includes a specific patent classification symbol It will be a complete set of documents. At this time, when a document set containing the patent classification symbol is obtained, documents containing the sub-patent classification symbol of the patent classification symbol should naturally also be included.

Meanwhile, the multidimensional analysis calculation execution result table generation module 402 of the present invention may store the calculation execution results in the same manner as the total amount analysis for the patent index that can be defined by other equations as well as the occupancy rate.

Hereinafter, an analysis index is introduced. (1) Revealed Technological Advantage, (2) Revealed Patent Advantage, (3) Concentration Ratio, and (4) (HHI, Herfindahl Index).

2. Technology level analysis indicators include 1) Patent Count Weighted by Citations, 2) Cites per patent, 3) Patent Impact Index (PII) 4) Current Impact Index (CII)

5) TS, Technology Strength, 6) Technology Cycle Time, 7) SL, Science Linkage, 8) Average Claims per Patent, , And 9) Family Size.

3. Cooperation Relationships and Knowledge Flow Analysis Indicators include 1) Number of Patents with co-inventors, 2) Salton's Index, 3) Workforce inflow rate, Brain Gain, Brain Drain), 4) Index for Knowledge Flow with Patent Citations,

It will be apparent to those skilled in the art that additional analytical indices that are more applicable than the above-mentioned patent indicators may be additionally introduced, all of which may be formulated, and that all formulas can be matched with a quality consciousness.

How to create a baseline ranking

The method of generating the ranking will be described in more detail. When the applicant A is obtained (confirmed), the document set of the applicant A in the first or second station can be obtained (confirmed), and the IPC or the like can be extracted from the document set. The IPC can count the number of documents by IPC at each step (from section to n dot subgroup) at each step, and the counted value is the total number of applications / It is possible to calculate the rank of the field of activity. (Generation of ranking information by the patent classification symbol level of the applicant A of the patent index standard has been described above.) The field with high ranking can be extracted by the calculated ranking, At least one extracted patent classification symbol such as IPC may be an upper ranking technology field. The method described above is a method of extracting rankings of the most favorable patent classification symbols for each stage of at least one corresponding patent classification symbol for the subject when given one subject (or extended attribute) (for example, , Inventors, agents, etc.) can also be created in the same way by the patent classification symbol level based on the document set containing their names.

In FIG. 19, the number of patent classification schemes such as IPC and the like in units of applicants for application / registration number, occupancy rate, concentration rate, and / or activity rate in each country such as Korea, USA, And that the reference ranking is generated for each level. In FIG. 19, it can be seen that H01L is ranked first in the IPC subclass level and H04N is ranked second in the case of Samsung Electronics based on the Korean application document.

The details will be described in detail as follows. 1) determining a set of at least one document that shares attributes (e.g., applicant A, or inventor A or agent B of applicant A, or a particular keyword common to a particular period of time, If you share at least one attribute such as quoting or quoting a document in a set of documents, you can create a set of partial documents that share that attribute in the entire document set through a query to the SQL query or search engine 2) extracting at least one corresponding patent classification symbol for each document constituting the document set, 3) extracting the extracted patent classification symbol by referring to the patent classification system, 4) All of the above patent classification symbols are classified by the level of the patent classification symbol (for example, in the case of IPC, n dot group), (5) counting individual documents including the stored patent classification symbols by roll-up by level, (6) performing calculation for each predetermined index by referring to the counting result 7) Calculating the ranking by considering the roll-up of each level of the patent classification symbol for each analysis index (the step of calculating the ranking, It is possible to extract rankings of the most favorable patent classification symbols for each step of at least one of the patent classification symbols in the first or second station. At this time, if the set of documents is limited to a certain attribute (eg applicant A (such as applicant A), applicant attributes such as last five years, inventor limitations such as inventor C, and / It may be possible to share the combined attributes (such as inventor C of applicant A). In the case of a set of documents by applicant, it may be possible to calculate 1) to 7) above for all applicants belonging to the first country It will be self-evident.

The multidimensional analysis operation result table generation module 402 of the present invention may generate at least one of the steps 1) to 7) to perform multidimensional analysis operation result table data.

It should be noted that in FIG. 19, a multi-filing reference IPC ranking is generated in the left column at the IPC subclass level. The numerical values shown in the embodiments of the present invention are obtained by processing the values input to the analysis module by a computational method and are variable depending on data supplement, deletion, modification, and so on. Therefore, There may be a difference. Therefore, it is to be understood that the inventive concept of the present invention is not limited to the numerical values shown in the tables and the like in the present specification, but is a structure, a configuration, an apparatus, a method, an information processing procedure, a system, and a utilization method for generating the numerical value. Hereinafter, the same applies.

19 is an exemplary diagram of an analysis result generated when drilling down to H01L. When drilling down, the analysis information value is generated only for the drill down. At this time, using AJAX technology, it is possible to generate only a numerical value for a part where a quick change (drilldown) occurs without reloading the entire page and provide it to a user. In the present invention, the user should be interpreted as the user computer 300 in relation to the system of the present invention. That is, in the system, 1) providing to the user is actually transmitted to the user's computer, 2) input by the user is actually receiving the input value of the user from the user's computer, Selection, and the like are all actually received from the user's computer by the user's designation, pre-selection, and selected information, it is appropriate that all expressions related to the user should be interpreted as described above. However, the reason for the description based on the user is due to the simplicity of description.

19 is an exemplary diagram of the analysis results generated when drilling down to H01L 21/00 and other subpublic classifications. It can be seen that the result of the analysis based on the total amount comes out by drilling down. Drilling down to the bottom of the IPC is possible, drilling down to the lowest level as long as there is data. This is also true for other patent classification codes.

FIG. 31 is an exemplary illustration of an application total amount analysis, including a drilldown on IPC H04B and their subclasses based on the US application entire document, as an example of a technology domain analysis using the patent classification symbol of the present invention. At this time, the analysis data of each cell includes all the data corresponding to the sub-patent classification symbol rolled up.

Multidimensional analysis operation result table for competition analysis

In order to perform the competition analysis, the multidimensional analysis calculation execution result table generation module 402 generates calculation result table data for multidimensional analysis for competition analysis. A description will be given of how the multidimensional analysis calculation execution result table generation module 402 generates the multidimensional analysis calculation execution result table data for competition analysis.

Competition from Applicant Perspective

Competition from the applicant's point of view can be defined as a conflict between the applicant A's country (first country) or another applicant B at least one other country (second country). On the other hand, the conflicts between applicants can be classified into 1) common in multiple application technology fields, 2) common in high concentration fields, 3) common in high activity fields, and in the field of application / It is preferable that the field is defined by at least one patent technical classification symbol for each level such as IPC for each level. For example, in the first country, when applicant A is filing a number of applications in H01L on the IPC subclass basis, 1) applicants (who have a high share) to H01L on the basis of IPC subclass in the first country, 2) 3) Applicants may be direct or potential competitors in the second country.) Meanwhile, Applicant A is a member of H01L When the concentration rate or the activity rate is high, the above 1), 2), 3) applicants of the first or second country may become competitors.

How to obtain competitive information from the applicant's point of view

Describe how to obtain competitive information from the applicant's point of view. A method of generating ranking information for each level of patent classification symbols of each analysis index (total amount, occupancy rate, concentration rate, activity rate, and the like) of each country of applicant A has been described above. For example, when H04B 7/02 is selected as a multi-filed subclass IPC in IPC 1 dot subclass (C5 level) based on Korean patent application document of Samsung Electronics Co., Ltd., the multidimensional analysis calculation execution result table generation module 402 Generates the following information.

2) Applicants who have a high concentration rate in H04B 7/02, 3) Applicants who have a high activity rate in IPC H04B 7/02 Applicant, or 4) applicants with high calculated values for other patent analysis indicators can be extracted. The above extraction method can be basically processed by an SQL query statement. In this case, when the operation table data for the multidimensional analysis is generated for each level of the patent classification symbol based on the analysis indexes such as the total amount, the occupancy rate, the concentration rate, and the activity rate, It is possible to obtain at least one applicant information having a high degree of competition relevant to the target and information such as the application / registration amount of the applicant's predetermined period of time by accessing the SQL query statement. If the multidimensional analysis operation result table data is not generated, the target information is obtained by using a relatively long and complicated SQL statement taking the following sequence.

The logic of the SQL statement may include 1) obtaining a specific level of patent classification code (for example, IPC H04B 7/02) and sub-patent classification codes of the patent classification code (the sub-patent classification codes are the patent classification code mast DB 203 (Which is described above), from the patent document master database 202 of the national unit / country integrated unit (in this case, the same document is repeated twice or more 2) obtaining date information such as applicant, filing date / registration date, etc. from the bibliographic information of the extracted document; 3) sorting on the basis of all applicants / And / or 4) a step of counting the application / registration amount in a predetermined period unit from the filing date / registration date information . Of course, it is needless to say that any two or more of the above 1) to 4) steps can be processed at once. Although the above SQL logic is described on the basis of the total amount, other patent analysis indexes such as the occupancy rate, concentration rate and activity rate can be processed in the same manner.

For example, in order to find an applicant having a high concentration rate, after the list of applicants is prepared after the step 2), the number of applications / registration number per applicant H04B 7/02 is updated with the number of all applications / (Which can be obtained as SQL in the DB 202), and then the sorting of the step 3) may be processed with the calculated concentration ratio. If the multidimensional analysis operation result table generation module 402 generates the multidimensional analysis operation result table data by the level of the patent classification symbol as described below, it can be processed with simple SQL. Such simple SQL would be extremely easy for a person skilled in the art.)

If obtaining the sub-patent classification symbol of the given patent classification symbol in step 1) can not be easily obtained as an extension (*,?, Etc.), there is a problem that much processing is required in the data processing of step 1). Therefore, in this case, the data generated through the roll-up operation based on the count information of the document related to the sub-patent classification symbol is more useful for the multidimensional analysis operation result table generation module 402 shown in Table 55 below.

[Table 56]

GC C5 C6 C7 C .. 01 02 03 04 05 H04B 7/02 H04B 7/04 H04B 7/06 0 0 3 17 12 H04B 7/02 H04B 7/04 H04B 7/06 0 0 3 17 12 H04B 7/02 H04B 7/04 H04B 7/08 3 One 5 4 3 H04B 7/02 H04B 7/04 H04B 7/08 3 One 5 4 3 H04B 7/02 H04B 7/04 3 2 2 21 24 H04B 7/02 H04B 7/04 3 2 2 21 24 H04B 7/02 H04B 7/04 6 3 10 42 39 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/10 0 0 0 0 0 H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 H04B 7/12 0 One 0 0 One H04B 7/02 9 3 20 18 24 H04B 7/02 9 3 20 18 24 H04B 7/02 9 3 20 18 24 H04B 7/02 15 7 30 60 64 H04B 7/02 50 40 30 30 30

When there is data such as the multidimensional analysis operation result table as shown in Table 56, a record having H04B 7/02 at the IPC 1 dot level (C5 column) is extracted, and the number of records is extracted as Appident field AppName group by, and rank it. In this way, it is possible to extract B and the like as a competitive applicant when the standard of H04B 7/02 of A is extracted, and the application / registration number per year / period of B is obtained by the applicant B It is possible to read out the processed value of the documents including the sub-patent classification codes of H04B 7/02 and H04B 7/02 among the documents and to provide them to the users.

The step of generating the multi-dimensional analysis operation result table generation module 402 generates at least one multi-dimensional analysis operation result table data for competition analysis is as follows. The multidimensional analysis operation result table generation module 402 generates multidimensional analysis operation result table data as shown in Table 56 for each type of competition set. When the multi-dimensional analysis operation result table generation module 402 generates table / data as shown in Table 65 with the occupancy rate, concentration rate, activity rate or other analysis index in the form of Table 65, the AppName, And the level of the patent classification symbol are equivalent, and the content of the numerical value per year / period may be the value of the occupancy rate, the concentration rate, the activity rate, or other analysis index.

1) Applicants who have filed multiple applications / multi-registrations in IPC H04B 7/02 based on application documents / registration documents in the first country; 2) Applicants with high concentration rate in H04B 7/02; 3) The IPC H04B 7/02 has been described in the same way on the basis of the second country on the basis of the obtained IPC H04B 7/02, (Based on the patent document data of the first country, the patent document data of the second country may be processed in the same manner as that processed by the first country). Of course, If there is a classifier, the symbols have a multi-level hierarchy like IPC, and the hierarchy can be distinguished by the number of dots and so on, so treat them as equivalent to IPC.)

33 is an exemplary diagram of a representative competitor analysis based on the total amount of the Korean standard Samsung Electronics Co., Ltd. among all the applicants in the DB held by the patent information system 1 of the present invention. The analysis of the representative competitor by the applicant is performed by obtaining the ranking and the application amount of the patent classification symbols of the IPC group level among the patent applications of the applicant, Based on the applicant's amount of application information, by a predetermined formula, and to provide applicants with high scores along with the rankings. It is preferable that the representative competitor analysis includes the number of applications / registration number of the representative competitors along with the year. Of course, when the application number / registration number is clicked, a document corresponding to the number of applications / registration number is obtained in the query formula embedded in the number of applications / registration number, and the document is transmitted to the simple analysis module 407, The module 407 provides simplified analysis of these documents. The numerical values in all the cells of the present invention are based on the query value, and a document for obtaining the numerical value from each query expression can be obtained. This applies to all cells below. If the value is a ratio value, the query is matched for each elemental value included in the formula for generating the ratio, so that the reference document can be obtained based on the query.

FIG. 33 shows an embodiment of the analysis of the competitive applicants according to the patent application technology classification codes of the total amount based on the Korean standard Samsung Electronics Co., Ltd. among all the applicants in the DB possessed by the patent information system (1) of the present invention. To obtain a competitive applicant for each specific technical classification code level and technical classification code, 1) obtain ranking information (reference number H01L, second place H04N, etc. in the above-mentioned FIG. 33) of reference patent classification symbols by the specific technical classification symbol level , 2) a specific country (selected by the user or automatically selected by the system (1) in Korea, the United States, Japan, Europe, etc.) by the reference patent classification symbol (for example, H01L in FIG. 33) , Applicants with high percentage of applications (based on application amount), high share (based on market share), high concentration (based on concentration ratio), high activity rate (based on activity rate) And 3) the application number / registration number of the applicant together with the extracted applicant information. At this time, if the drill-down is performed, the above-mentioned 1) to 3) are performed on the sub-patent classification code of the drilled-down patent classification symbol with one applicant's document to provide information on the competing applicant. The drill-down can be drilled down to the lowest patent classification symbol. On the other hand, the competitor may be carried out in the same manner not only for the IPC but also for the USPC and the FT. However, the selection of the country will be limited. (Since USPC is a US patent classification code, The applicant can not be obtained.) The competitive applicant can be created based on the total amount of the applicant's application, but it can also be created on the basis of the total amount of the applicant's registration.

On the other hand, if the reference patent classification symbol is generated on the basis of the patent document of the specific applicant in the first country, and the applicant who is highly competitive, highly occupied, highly motivated, And may be extracted from the second country. The selection of the first station and the second station may be selected by the user, but may be selected by the system 1 as the default value.

On the other hand, it is possible to analyze the entry competitor applicant, and the generation of the analysis information for the entry competitor will be able to be processed only for documents of a predetermined recent period (for example, within the last 7 years). That is, 1) information on the criteria of the patent classification symbol per the patent classification symbol level is generated based on the applicant's document in the first country registered / registered in the latest period, and i ) Recent period, or ii) on the basis of the entire period, can be extracted from the first country or the second country. 2) Generate information on the criteria of the patent classification symbol per the patent classification level based on the documents of the applicant in the first country registered / registered in the whole period, Based on the above, it is possible that an applicant who is highly competitive, highly occupied or highly competitive can be extracted from the first country or the second country.

Competition from an inventor's perspective

In the above description, how the multidimensional analysis calculation execution result table generation module 402 generates data in order to obtain competition information from the applicant's perspective has been described. The inventor's view is to compile the table / data for each analysis indicator such as the total amount, the occupancy rate, the concentration rate, the activity rate, etc. by the method of processing the competition from the viewpoint of the applicant Data is accessed on the basis of the applicant) (a set of documents is created for each inventor, and the table / data is generated in an equivalent manner by the set of documents for each inventor, and then the data is accessed by the inventor ). Table 57 below shows an example of such data.

Inventor analysis multi-dimensional analysis operation result table

The data generated by the multi-dimensional analysis operation result table generation module 402 based on the inventor may be two sequences. The first is a method of generating data by placing an inventor in the lower part of one applicant and processing a set of documents of the applicant's inventor (that is, the applicant is usually a large corporation or organization, It is possible to generate data for each set of documents satisfying the above conditions, such as applicant A AND inventor 1, applicant A AND inventor 2, and the like. ), And second, data is generated in such a manner that only the inventor is treated independently (in the form in which the inventor enters the applicant's place) and the set of documents of the inventor is processed. One example of the data format generated by the multi-dimensional analysis operation result table generation module 402 for the latter is shown in Table 57 below because the former will be obvious.

[Table 57]

IAppName Inventor C3 C4 C5 C6 00 01 02 03 04 05 A a H04N H04N 5/00 H04N 5/64 H04N 5/655 0 0 0 One 0 0 A a H04N H04N 5/00 H04N 5/64 H04N 5/655 0 0 0 One 0 0 A a H04N H04N 5/00 H04N 5/64 H04N 5/655 0 0 0 One 0 0 A a H04N H04N 5/00 H04N 5/64 One 2 0 0 0 0 A a H04N H04N 5/00 H04N 5/64 One 2 0 0 0 0 A a H04N H04N 5/00 H04N 5/64 One 2 0 0 0 0 A a H04N H04N 5/00 H04N 5/64 One 2 0 One 0 0 A a H04N H04N 5/00 H04N 5/72 0 One 0 0 0 0 A a H04N H04N 5/00 H04N 5/72 0 One 0 0 0 0 A a H04N H04N 5/00 H04N 5/72 0 One 0 0 0 0 A a H04N H04N 5/00 H04N 5/72 0 One 0 0 0 0 A a H04N H04N 5/00 H04N 5/74 0 One 0 One One 0 A a H04N H04N 5/00 H04N 5/74 0 One 0 One One 0 A a H04N H04N 5/00 H04N 5/74 0 One 0 One One 0 A a H04N H04N 5/00 H04N 5/74 0 One 0 One One 0 A a H04N H04N 5/00 One 4 0 2 One 0 A a H04N One 4 0 2 One 0 A b B z

When the multidimensional analysis operation result table generation module 402 has the total amount information as described above, a method of generating data such as occupancy rate, concentration rate, activity rate, and the like for each individual inventor unit of a specific applicant is equal to As described in the section describing how to generate information. Of course, when the data as shown in Table 57 has all the inventor's information for all applicants, the multidimensional analysis calculation execution result table generation module 402 may calculate the total amount information in units of individual inventors , Data on the values of the patent analysis index such as the occupancy rate, the concentration rate, and the activity rate can be generated.

FIG. 35 is an exemplary diagram of an analysis result of multi-filing inventors of Samsung Electronics Co., Ltd. based on the total amount of Korean patent applications among all the applicants in the DB held by the patent information system 1 of the present invention. The inventor's analysis of the patent classification code level by area includes: 1) generating a patent classification symbol reference rank for each patent classification symbol level of a specific applicant; 2) calculating, by the patent classification symbol standard ranking, i) (Ii) calculating the rankings of the inventors (e.g., the first to tenth places) of the inventors according to the year integration, (3) a step of realizing any one of the steps of (3) To the user. On the other hand, the analysis of the inventor by category according to the level of the patent classification symbol may be performed on the basis of the application document or the registration document without specifying the applicant on a country basis. In this case, 1) generating a patent classification symbol reference ranking for each patent classification symbol level on the basis of the entire object document, 2) calculating information on the maximum number of applicants for each year by the patent classification symbol reference ranking, ii) (1) to (10), and (3) a step of calculating the inventor's analysis information for each year, and (3) It breaks down.

36 is an example of an analysis result of multi-application multi-applicant inventors of Samsung Electronics Co., Ltd. based on the total amount of Korean patent applications among all applicants in the DB possessed by the patent information system (1) of the present invention. The inventor's individual analysis is performed by the inventor's individual analysis information for all inventors (inventors who have filed a patent application) related to a specific applicant at a country level. The inventor's individual analysis information is obtained by 1) obtaining all application documents or registration documents related to the applicant 2) extracting the inventor from the obtained document, 3) counting the number of documents per inventor according to a predetermined policy, 4) generating a ranking for each inventor, and 5) And generates the number of applications / registration number information for each star / specific period and provides it to the user. On the other hand, if it does not relate to a specific applicant, obtain all patent documents of the country unit and execute the above 2) ~ 5). Such documents may be subject to documents for a predetermined period, such as the last seven years. This is essentially a matter of creating a document set, and the creation of the document set may be any one of the methods of generating the document set referred to elsewhere in this specification.

Citation analysis multi-dimensional analysis operation result table

In order to perform citation analysis, the multi-dimensional analysis operation result table generation module 402 generates multi-dimensional analysis operation result table data for citation analysis. As to how the multidimensional analysis operation execution result table generation module 402 generates the multidimensional analysis operation execution result table data for citation analysis through a certain process, U.S. Patent Application No. 09 / 802,847 (filed by Samsung Electronics Co., LTD (KR), entitled &quot; User request processing method and apparatus using upstream channel in interactive multimedia contents service, filed March 12, 2001).

The bibliography of the above-mentioned U.S. Patent Application No. 09 / 802,847 is as shown in Table 42.

[Table 42]

Application number Filing date Registration Number Registration date Citation Information (US Patent) IPC USPC 09 / 802,847 March 12, 2001 7,302,464 November 27, 2007 5680322
5805804
6044397
6130898
6317131
6611262
6631403
6654761
6654931
6697869
2002/0026642 G06F 15/16

H04N 7/16 709/203

709/217

725/135

At this time, the multi-dimensional analysis operation result table generation module 402 obtains the following cited parent-child data from the bibliography information of 09 / 802,847 regarding citation. A child is his / her own document number, and a patent is a reference number that he / she is citing (in this case, his / her document number may be an application number or a registration number, that is, However, since the parent may have a publication number or an application number, it is preferable to uniformize the application number, which is a common number of all the documents. However, it is also possible to unify the application number with the registration number, If the complexity of the processing is not a concern, it is also possible to mix several numbers if there is no problem in the specification of the document.) In the present specification, the application numbers are unified for ease of explanation. -child data.

[Table 43]

child-country child (application number) parent parent-country US 09 / 802,847 5680322 US US 09 / 802,847 5805804 US US 09 / 802,847 6044397 US US 09 / 802,847 6130898 US US 09 / 802,847 6317131 US US 09 / 802,847 6611262 US US 09 / 802,847 6631403 US US 09 / 802,847 6654761 US US 09 / 802,847 6654931 US US 09 / 802,847 6697869 US US 09 / 802,847 2002/0026642 US

As shown in Table 43, the cited parent-child data unified with the application number is generated from the obtained parent-child data as shown in Table 44 below. Each application number is obtained from the parent's registration number and public number (the application number is acceptable, all number information has a document type, A registration, A1 disclosure, etc.) (US patent application number or application serial number The data shown in the following table is referred to as a unified number reference unit reference parent-child data. On the other hand, the national columns of the documents in the child and parent columns refer to the country of origin from which the document concerned is related (filed or registered). Such a country can be applied not only to the United States but also to various countries. In the following, the national column is omitted for convenience of explanation, but it should be understood that there is a national column. On the other hand, the notation of the application number may be any notation such as "two digits + / 6 digits" or "year + 6 digits &quot;.

[Table 44]

child (application number) parent (application number) 09 / 802,847 08 / 451,470 09 / 802,847 08 / 816,207 09 / 802,847 09 / 055,929 09 / 802,847 08 / 969,965 09 / 802,847 09 / 113,748 09 / 802,847 09 / 152,003 09 / 802,847 09 / 309,895 09 / 802,847 09 / 124,474 09 / 802,847 09 / 236,462 09 / 802,847 09 / 138,782 09 / 802,847 09/736393

Although the data processing as shown in Table 44 is generated based on a given document, the data processing as described above may be processed in units of individual documents belonging to the given document set in at least one document set unit, Table 45 summarizes these. At this time, it is possible to allow duplication of parent (application number) in the unit of individual denominations (ie, if documents 1 and 2 belonging to one document set contain document 3 in the reference information, It is possible to create a parent-child data citing a uniform number base unit for every individual document belonging to the set of documents. At this time, the most representative document set can be a set of US open application documents and a set of US registered patent documents. It will be obvious that a set of documents created by the user (outputting a search result after inputting a specific search expression) and the computerized system 1 implementing the present invention can be automatically targeted.

[Table 45]

child (application number) parent (application number) 09 / 802,847 08 / 451,470 09 / 802,847 08 / 816,207 09 / 802,847 09 / 055,929 09 / 802,847 08 / 969,965 09 / 802,847 09 / 113,748 09 / 802,847 09 / 152,003 09 / 802,847 09 / 309,895 09 / 802,847 09 / 124,474 09 / 802,847 09 / 236,462 09 / 802,847 09 / 138,782 09 / 802,847 09/736393

In addition, if the patent document master DB 202 is searched for all the documents included in the parent-child data citing the unified number reference unit, the data with the bibliography added to each document can be generated as shown in Table 46 below . The addition of a bibliography may be any one or more of the elements constituting the bibliography. Table 46 below shows data added with the application date and the registration date among the bibliographies corresponding to only one application number.

[Table 46]

child registration date child filing date child application number parent application number parent Filing Date parent Registration Date November 27, 2007 March 12, 2001 09 / 802,847 08 / 451,470 May 26, 1995 October 21, 1997 November 27, 2007 March 12, 2001 09 / 802,847 08 / 816,207 March 12, 1997 September 8, 1998 November 27, 2007 March 12, 2001 09 / 802,847 09 / 055,929 April 7, 1998 March 28, 2000 November 27, 2007 March 12, 2001 09 / 802,847 08 / 969,965 November 25, 1997 October 10, 2000 November 27, 2007 March 12, 2001 09 / 802,847 09 / 113,748 July 10, 1998 November 13, 2001 November 27, 2007 March 12, 2001 09 / 802,847 09 / 152,003 September 11, 1998 August 26, 2003 November 27, 2007 March 12, 2001 09 / 802,847 09 / 309,895 May 11, 1999 October 7, 2003 November 27, 2007 March 12, 2001 09 / 802,847 09 / 124,474 July 29, 1998 November 25, 2003 November 27, 2007 March 12, 2001 09 / 802,847 09 / 236,462 January 25, 1999 November 25, 2003 November 27, 2007 March 12, 2001 09 / 802,847 09 / 138,782 August 24, 1998 February 24, 2004 November 27, 2007 March 12, 2001 09 / 802,847 09/736393 December 15, 2000

Count information (number of claims, number of drawings, number of families, number of inventors, number of applicants, and various counting information may be additionally added for each count field in the data as shown in the above Table 46) It is preferable to generate one table for items corresponding to only one item.

Then, the above-mentioned data may be generated by using a field value in which two or more pieces of data can be associated with each application number. These may be applicants, inventors, agents, patent classification codes, and the like. The patent classification codes will be described later. If two or more data correspond to each application number, you can put all the required field values into one table, but you have to face the problem of large table size. For example, in U.S. Patent Application Serial No. 09 / 309,895, (New York, NY) and Sun Micro Systems (Palo Alto, Calif.), Where the applicant field is added to the line shown in Table 47 below with respect to this application number, .

[Table 47]

child registration date child filing date child application number parent application number parent Filing Date parent Registration Date November 27, 2007 March 12, 2001 09 / 802,847 09 / 309,895 May 11, 1999 October 7, 2003

[Table 48]

child registration date child filing date child application number parent application number parent Filing Date parent Registration Date Applicant November 27, 2007 March 12, 2001 09 / 802,847 09 / 309,895 May 11, 1999 October 7, 2003 AT'T Corp November 27, 2007 March 12, 2001 09 / 802,847 09 / 309,895 May 11, 1999 October 7, 2003 Sun Micro Systems

Thus, if there are two applicants, the number of rows will be doubled. On the other hand, if the applicant and the inventor are present in one table, 12 rows are required if the inventor has six applicants. (2 * 6 = 12) At this time, (Of course, there is a problem that duplication occurs in all the data except for the table.) (Of course, when there is duplication, when the analysis result (for example, the applicant for each year and the inventor of the specific applicant for each year) There is a problem in that the size of the recording medium is large.

For reference, it is desirable that the information processing based on time / date is processed as follows when considering roll-up. For example, in the case of March 12, 2001, it is desirable to perform the roll-up operation, such as March 12, 2001, March 2001, 2001, and 2001. However, in the present specification, this is omitted because of the problem of representing (representing a large number of various fields on a limited width screen), which will be obvious to those skilled in the art.

On the other hand, with respect to the patent classification symbol, the processing of the data will be described from the viewpoint of inclusion of the sub-classification symbol. First, as shown in the following Table 49, data processing for 09 / 802,847 corresponding to child is explained, and the same applies to documents belonging to the parent. 09 / 802,847 has IPCs G06F 15/16 and H04N 7/16, USPC 709/203; 709/217; 709/231; 725/135.

[Table 49]

Application March 2001 Application year 2001 n dot subgroup (C (n + 4)) 1 dot subgroup (C5) IPC group (C4) IPC subclass (C3) IPC Class (C2) IPC Section (C1) child application number One One G06F 15/16 G06F 15/00 G06F G06 G 09 / 802,847 One One H04N7 / 16 H04N 7/00 H04N H04 H 09 / 802,847

In Table 49, C3 denotes the IPC subclass level, and the others are the same as above. One field in 2001 refers to one in 2001, and there is one in March 2001, which means that there is one per month in month, and the above is a common notation in data structure considering roll-up . All of the above G06F 15/16 and H04N 7/16 belong to the 1dot subgroup, but if H04N 7/169 is included in the document 09 / 802,847, Table 49 above would be as shown in Table 50 below. Which was arbitrarily introduced for the purpose of explaining the inventive idea)

[Table 50]

Application March 2001 Application year 2001 3 dot (C7) 2 dot (C6) 1 dot subgroup (C5) IPC group (C4) IPC subclass (C3) IPC Class (C2) IPC Section (C1) child application number One One 09 / 802,847 One One H04N7 / 169 H04N7 / 167 H04N7 / 16 H04N 7/00 H04N H04 H 09 / 802,847

The above data may be processed as shown in Table 51 from the viewpoint of inclusion of the USPC in the sub-classification code.

[Table 51]

Application March 2001 Application year 2001 USPC n dot level (C (n + 4)) USPC 1 dot level (C5) USPC no dot level (C4)  USPC class (C3)  Meta class (C2) Meta superclass (C1) child application number One One 709/203 709/201 709 09 / 802,847 One One 709/217 709 09 / 802,847 One One 709/231 709/230 709 09 / 802,847 One One 725/135 725 09 / 802,847

USPC 709/203 (title information: Client / server) is in the top 709/201 (title information: DISTRIBUTED DATA PROCESSING), which belongs to Class 709. Other things are the same. All of the above USPCs and the like belong to the 1 dot or no dot level, but if the document No. 09 / 802,847 contains 725/45, Table 51 will be as shown in the following Table 52. Which is arbitrarily introduced for the purpose of the present invention)

[Table 52]

Application March 2001 Application year 2001 4 dot (C8) 3 dot (C7) 2 dot (C6) USPC 1 dot level (C5) USPC no dot level (C4)  USPC class (C3) Meta class
(C2) Meta superclass (C1) child application number One One One One 725/45 725/44 725/39 725/38 725/37 725 09 / 802,847

 IPC is assigned H04L 12/56, H04L 12/28, USPC 370/395, and 370/235. (The above IPC may be different from that granted by the Korean Intellectual Property Office, Is preferably treated with the IPC granted by the USPTO.)

At this time, since one line is created for each patent classification symbol, IPC and USPC can be created in one table, but it is preferable not to create one. This is especially true if there are a plurality of applicants, such as a plurality of inventors, etc., but there are essentially several parents in one child. That is, there are k parents in one child, m IPCs in one child, n USPCs, and there will be IPCs and USPCs in each k each parent document. If you do, you will need a huge number of lines, at which time a lot of data will be duplicated. This is even more so, especially if the size of the document set is large (eg, a full set of US patents). Therefore, it is desirable to separate IPC and USPC. In relation to the patent classification, the table type is: 1) the type that is processed only for the child patent data in the child-parent data, 2) the type processed for the parent patent data only in the child-parent data, and 3) There may be a type of child-parent data that has been processed only for the patent classification symbol for both the child and parent sides. 3), there is a problem that the number of rows increases, but the possibility of table joining is reduced.

That is, the multi-dimensional analysis operation result table generation module 402 generates unit citation parent-child data for individual documents belonging to a given / set document set, and for each unit citation parent-child data, And / or a bibliography of individual documents belonging to the parent, and generates the multidimensional analysis operation result table data in which the obtained bibliography is the field contents.

Examples of the set of given / preset documents are 1) a complete set of application documents or registration documents in a particular country, 2) a set of documents in one or more of the specific IPC / USPCs, 3) a set of documents by specific applicant, 4) And 5) a complete set of all national application documents or all registration documents.

At this time, if the data as shown in Table 62 below are generated for the entire document set in the above 1) country unit, the following utility may be obtained. It will be appreciated that bibliographic data may be stored in each application number / registration number constituting the data, and any one or more of the bibliographic data may be combined and generated as described above.

[Table 53]

child (application number) parent (application number) 09 / 802,847 08 / 451,470 09 / 802,847 08 / 816,207 09 / 802,847 09 / 055,929 09 / 802,847 08 / 969,965 09 / 802,847 09 / 113,748 09 / 802,847 09 / 152,003 09 / 802,847 09 / 309,895 09 / 802,847 09 / 124,474 09 / 802,847 09 / 236,462 09 / 802,847 09 / 138,782 09 / 802,847 09/736393

The method of generating the above data is to put all application / registration documents in the child column, and insert the parent document number which is citing the document for each document into the parent. If there is no parent value, a null value is inserted. In this case, there is only data in the child column, and there is no parent document corresponding to the child. In this case, the child column will contain all the document numbers.

For simplicity, let's introduce the simplest model. (1), (2), (3), and (4) contain one citation information, and the order of citation is 1 -> document 2 -> document 3 -> document 4 , Document 3 refers to document 2, and document 4 refers to document 3). If document 2 is the base (document 2 is in the child column), document 1 is in the parent column of document 2, document 3 is the base (document 3 is in the child column), document 2 parent "column, and based on document 4, document 3 is in the parent column. That is, documents 1 through 4 are all in the child column, and at least the documents 1 through 3 are in the parent column. The depth of the quotation (when document 4 is used, document 1 has a depth of 3, document 2 has a depth of 2, document 3 has a depth of 1), backward citation, The document set generated from the citation information included in document 3) will be document 2 and document 1 from depth 2, and document 4 will be at depth 1 for forward citation. In this case, since documents 1 to 4 are all in child, it is possible to grasp the citation document with depth 1 in the parent column based on any document, find the cited document number again in the child, find the parent of the found child If you do, you will be able to find a citation document with a citation depth of two. These models are described in Table 54 as follows.

[Table 54]

child (application number) parent (application number) 2 One 3 2 4 3

In the above, one document has been explained, but it will be obvious that this also applies to the document set. This is also described by a model as shown in Table 55 below. (The contents in the line in the following Table 55 are not cell data of the operation result table of the actual multi-dimensional circle analysis but the cell data of several lines of the concept are one set) And should be interpreted in the sense of the present invention for convenience of explanation.)

[Table 55]

child (application number) parent (application number) Document set 2 Document set 1 (set generated by all parent of document set 2) Document set 3 (share specific attribute 3) Document set 2 (set generated by all parent of document set 3) Document set 4 (set of all children of document set 3) Document set 3 (set generated by all parent of document set 4)

In Table 55 above, document sets 1 through 4 are all a subset of the entire set of child columns, and document sets 1 through 3 are all a subset of the entire set of parent columns. Subsetting means that all documents belonging to the set belong to the entire set, so they can be identified and extracted from the entire set. The attributes of the document set may be arbitrary attributes, but may be any of the following: 1) the name of the applicant; 2) the name of the inventor; 3) IPC or USPC at each stage; 4) 7) the status of the document (filing, registration, rejection, etc.), or 8) a combination of at least one of these may coincide.

For example, suppose that the document set belonging to the applicant Samsung Electronics, IPC H01L, and all registered documents registered in the United States are set to document set 3, a set of documents consisting of all quoted documents of the depth 1 2 will be extracted. On the other hand, it is possible to obtain the document set 4 consisting of all the citation motions of the backward reference depth 1 related to the document set 3 and the document set 1 of the front reference depth 2 related to the document set 3. At this time, the document set 1, the document set 2, and the document set 4 can be analyzed. 2) front / rear patent applicants; 3) front / rear patent applicants; 4) front / rear patent IPC / USPCs (each of the patent classification schemes); Step-by-step, drill-down has already been explained), and 5) you will find individual documents for front / rear multiple. On the other hand, in case of 1) above, the total amount may be divided by year / predetermined period, and the increase / decrease rate may be known based on various numbers. In case of 2) above, the amount of front / rear quotation for each applicant for multiple applicants will be divided by the total amount and / or year / period, which will be the same in 3) to 5). If the multiple applicants for the backward quotation of 2) are selected based on the document set 3 as the analysis target, information on the applicant who has filed the US registered document belonging to the Samsung Electronics IPC H01L will be known. This set of later applicants may include Samsung Electronics, which is a self-citation, and can be used to divide the numerical value of the self-citations into different numerical values, or to perform various analyzes of self-citations by pre- will be. In addition, if the applicant of the latter application is viewed by ranking, it is possible to know which applicant is quoting a lot of documents belonging to the Samsung Electronics IPC H01L. If the user clicks the link, a related document (backward reference document) is obtained from the patent document master DB 202, and the user inputs the related document (backward reference document) You will be able to provide it.

On the other hand, when one document set is given (e. G. Document set 3), the document set 2 which is the parent of the document set 3 is searched for, and in the document set 2, , And a document set 2 '' 'that satisfies a periodic condition such as a document is applied to the document set 2' '', and performs the at least one predetermined analysis by using the document set 2 '' 'as an analysis target document set, It is natural to be able to do.

Four types of document set for citation analysis

4 types

On the other hand, when citation analysis is concerned, there will be four kinds of documents to be analyzed. When a set of reference documents is given, the following four set of citation analysis documents are determined with reference to the citation.

FIG. 38 shows that there may be a set of documents to be subjected to the four kinds of analysis regarding the citation analysis for the reference document set. FIG. 38 shows an example in which "the third party patent (the third party) quoted by the company", "the third party patent (the third party) quoting the company", the third party's patent (the third party) quoted by the third party " Quot; patent (quot; - > ta) ", respectively, each of which is sequentially corresponded to the first to fourth.

First, it is a set of backward cited documents cited by the individual reference documents contained in the reference document set. The total set of forward quoted documents may be determined as a union of all documents corresponding to the reference cited included in each of the reference documents. (In this case, the union may be a union obtained by removing duplicates, For example, if a document # 1 and a document # 2 contain a document a as citation information, a may be an important document quoted twice in the counting position, so duplicate (There are duplicate records in the parent in the multi-dimensional analysis operation result table, that is, there are two records related to the parent a due to # 1 and # 2). On the other hand, It would be desirable to simply remove redundancy (by performing a union operation) in the presence of a (applicant, date, and other bibliography). In the analysis of the present invention, , It is more reasonable to allow duplication.)

Second, it is a set of forward quotation documents citing individual reference documents included in the reference document set. The method for obtaining the total document set of backward quotation is as follows: 1) a document number citing the individual reference document is searched and summed, or 2) a reference document number is generated in the child-parent table generated for the entire patent document, , Find the child document number (which may be more than one) corresponding to the reference document number in the parent column in the child column, and sum them together. 2) is more preferred.

Third, it is a set of reference documents related to the entire backward citation. This set of documents is a set of documents consisting of all the reference documents that constitute the reference document set and all the reference documents that have been quoted by other documents. This document set is a document set consisting of the parent of the second document set.

Fourth, it is a set of reference documents related to the entire forward citation. This set of documents is a set of documents consisting of reference documents containing cited information among all the reference documents constituting the reference document set. This document set is a document set consisting of the children of the first document set.

The uppermost tab in FIG. 39 includes 1) "citation analysis for a total set of patent documents" 2) "precise analysis of citation by technology area" 3) "citation analysis by inventor" 4) "citation analysis for individual documents "All of which are related to the establishment of a set of reference documents. The above 1) "citation analysis of the entire set of patent documents" means a complete set of patent documents as a set of reference documents determined by any of the document selection methods related to the applicant or the determination of the document set of the present invention. The above 2) "Precise Analysis of Citation by Technology Area" is a process for classifying the entire set of patent documents into a set of documents classified by patent classification symbol sorted by the patent classification symbol reference rank by level of the patent classification symbol (IPC, USPC, It means that it is a set. The above-mentioned 3) "citation analysis by inventor" means that the inventor is extracted from the entire set of the patent documents, and then the ranking is calculated, and the set of documents classified by the inventor having the highest ranking is used as the reference document set. The above-mentioned 4) "Analysis of citation by individual document for multiple documents" is to examine the number of documents cited / cited for all documents included in the entire set of documents, extract documents with high citation / pile ranking, It means that it is a set of reference documents.

The multi-dimensional analysis operation result table generation module 402

When the reference document set is determined, the multidimensional analysis operation result table generation module 402 generates a document set of 1) an entire forward reference document as the document set to be analyzed for the reference corresponding to the reference document set, 2) And / or 4) a set of related reference documents for the entire backward reference, and a multidimensional analysis described herein for any one or more of the document sets 1) - 4) And stores the result table.

Results of citation analysis

 The analytical module of the present invention may include a set of documents identified by 1) an entire forward citation, 2) a set of full backward references, 3) a set of reference documents related to an entire forward citation, and / or 4) The analysis result is obtained by a predetermined analysis formula on the operation result table for the multi-dimensional analysis by the subject of the analysis. Meanwhile, the analysis module may perform various analyzes described in the specification of the present invention for any one or more of the document sets 1) to 4) other than the multidimensional analysis operation result table, but in this case, There is a problem that the result of the multidimensional operation can not be obtained or the computer resources are consumed in order to obtain the result of the multidimensional operation.

The analysis module accesses the multidimensional analysis operation result table generated for each document set of 1) to 4), and extracts a desired citation analysis result in a predetermined analysis formula. The results of the analysis of the desired citation may include the following.

I) ranking of patent categorizers by at least one level in at least one or more patent categorization systems by applicant, iii) multiple inventor ranking, iv) document appearance frequency, The reference rankings and / or the numerical value data rolled up into a time dimension or another dimension for each of them, and / or information on the document number corresponding to each of them. That is, for the citation analysis target document set of 1) to 4) described above, a ranking value for each field included in the bibliographic data and / or a value obtained by roll- It will be able to generate information on the document number to obtain the individual document itself.

(Ii) the applicant belonging to the patent classification code per level, (iii) the patent classification by level, (iii) the patent classification by level, Iv) numerical value data of the inquiry document belonging to the level-specific patent classification symbol and / or numerical value data which is rolled up into a time dimension or another dimension for each of them, and / or a document corresponding to each of them Information about the number can be extracted.

That is, for the citation analysis target document set of 1) to 4) described above, a ranking value for each field included in the bibliographic data and / or a value obtained by roll- It will be able to generate information on the document number to obtain the individual document itself.

Third, based on the attributes of the applicant (applicant's type (enterprise, university, individual, etc.), financial property, corporate evaluation index, etc.), i) type of applicant, ii) , iii) rankings of corporate financial evaluation elements such as an annual average profit rate, and / or numerical value data rolled up into a time dimension or another dimension for each of these, and / or information about a document number corresponding to each of them . That is, for the citation analysis target document set of 1) to 4) described above, a ranking value for each field included in the bibliographic data and / or a value obtained by roll- It will be able to generate information on the document number to obtain the individual document itself.

Fourth, when there are numerical value data in the above first to third data, there are data on the change values such as the rate of increase / decrease, the rate of increase / decrease.

The multidimensional analysis operation result table generated by the multidimensional analysis operation execution result table generation module 402 (cited applicant reference) for citation analysis is mainly used for a patent document cited from the applicant's own document . On the other hand, it will be obvious that the same multidimensional analysis calculation execution result table can be generated for the above-mentioned examiner's citation document. That is, if the examiner refers to a document cited in the examination process for a specific application document (a child document) during the examination process as a parent document of the application document, the multidimensional analysis operation execution result table generation module 402 ) And the result of the operation (table of examiner quotation) citation analysis multi-dimensional analysis operation result table and the method of using the result are completely equivalent to those based on the applicant's citation document. Meanwhile, for an integrated citation analysis, it is possible to generate a parent-combined document by performing a union calculation of the applicant's citation document and the examiner's citation document (parent 2) based on a single application document, If the same process as that based on the applicant's citation document is performed on the integrated citation document set in the above, an equivalent result is obtained, and a method of using the equivalent result is also equivalent. The description in this paragraph is self-explanatory to those skilled in the art with reference to the description based on the document of the applicant's citation, so it is not separately redundantly described.

This will be described in more detail with reference to the drawings.

38 shows a set of backward quoted documents of the present invention when all the US patent applications of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system 1 of the present invention are set as reference documents, , Which is an example of the results of yearly analysis of the total amount of citations. 38 shows that the total amount tab is selected. In the illustrated embodiment of the present invention, various tabs are displayed in the exemplary embodiment of the present invention. The types of the tabs are basically 1) criteria for the property of the document set at the time of establishing the document set such as the standard of the application document or the document of the registration document, 2) the standard of the application / registration total amount, 3) standards related to the determination of countries such as the Republic of Korea, the United States of America, Japan, and Europe, and the criteria for the purpose and subject of analysis, such as applicants, inventors, 6) criteria for the selection of patent classification symbols such as IPC, USPC, FT, FI, ECLA, etc. and the level of each selected patent classification symbol. have.

39 is a diagram showing a set of backward quoted documents according to the present invention when all the US patent applications of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system 1 of the present invention are set as reference documents, , Which is an exemplary embodiment of the analysis result of the applicant of the present invention by year.

FIG. 40 shows a set of backward quoted documents of the present invention when all the US patent applications of Samsung Electronics Co., Ltd. among all the applicants in the database held by the patent information system 1 of the present invention are set as reference documents, (IPC main group level) according to an embodiment of the present invention. At this time, the IPCs can be drilled down. At this time, the citation analysis result of the sub-patent classification related to the drill-down is extracted and provided to the user. Figure 41 shows that.

41 is a diagram showing a set of backward quoted documents according to the present invention when all the US patent applications of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system 1 of the present invention are set as reference documents, (IPC main group level), which is a multi-patent classification code (IPC main group level).

FIG. 42 shows a set of backward quoted documents of the present invention when all the US patent applications of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system 1 of the present invention are set as a reference document set, , Which is an example of an analysis result of the inventor by year for each year.

FIG. 43 is a diagram showing a set of documents of the present invention, which is a set of reference documents of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system (1) , Which is an exemplary embodiment of the analysis results of the most cited applicants by year.

FIG. 44 is a diagram showing a set of documents of the present invention which are forward references of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system 1 of the present invention is set as a reference document set, , Which is an example of the results of yearly analysis of the most cited inventors.

FIG. 45 is a diagram showing a set of documents of the present invention, which is a set of reference documents of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. is set as a reference document set among all applicants in the DB held by the patent information system 1 of the present invention, , Which is an exemplary embodiment of the analysis results of years when drilling down to the IPC among the analysis of the most cited technologies per IPC main group.

46 shows a set of forward quotations of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system 1 of the present invention is set as a reference document set, (No dot, subclass subclass). The results of the analysis are summarized as follows.

47 shows a set of reference documents related to the entire rear citation of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system 1 of the present invention is set as a reference document set, And a chart generated by the chart generation module 406-2 of the reporting module 406 of the present invention with respect to the analysis result. The reporting module 406 of the present invention includes a table generating module 406-1 for generating a table, a chart generating module 406-2 for generating a chart, a graph generating module 406-3 for generating a graph, And a report generation module 406-4 for generating a report.

48 shows a set of reference documents related to the entire rear citation of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. among all the applicants in the DB held by the patent information system 1 of the present invention is set as a reference document set, , The results of the analysis on the total number of citations by the inventors who received a lot of citations and the results of the simple analysis on the set of documents corresponding to the number when the specific number shown in the analysis result is clicked A simple analysis module 407 for providing the number of applications / registration by year, the number of applications / registration by year of the most inventors, and the number of applications / registration numbers (including drill-down) by the maximum technology field (IPC, USPC, FT) FIG. 8 is an exemplary diagram of a list of generated documents. FIG.

49 is an exemplary diagram showing a drill-down function according to the maximum technology field (IPC, USPC, FT) in the simplified analysis module 407 of the present invention.

FIG. 50 is a diagram showing the relationship between the backward and forward directions of the present invention when all of the US application documents of Samsung Electronics Co., Ltd. among all the applicants in the DB held by the patent information system 1 of the present invention are documents of multi- ) Is an example of an analysis result of the total amount of citations, which is a set of documents to be analyzed, by year.

Patent analysis result for multi-dimensional analysis

Next, we will explain the technical analysis of the utilization of patent technology classification. Patent Technology Classification Utilization Analysis refers to the analysis of a set of documents determined through patent classification symbols among the document sets of the present invention. Patent Classification Techniques Generated by Patent Classification Classes The types of document classifications used for analysis include 1) documents containing at least one patent classification code selected from the types of specific patent classification codes (IPC, USPC, FT, ECLA, etc.) (Including various operations such as OR operation, AND operation and NOT operation) when two or more patent classification codes are intervened, (2) a set of documents generated including two or more kinds of patent classification symbols IPC AND USPC). All of these can be 1) individual country units, 2) national integrated units, and national redundancy should be eliminated (as a result of union), 3) whether to target only the main patent classification code, Whether or not to include it will depend on the pre-established policy or user's choice, all of which are related to the confirmation of the document set.

The information that can be extracted from the patent technology classification utilization technology analysis (hereinafter, referred to as the technology analysis) is a patent document classification analysis technique document analysis set (hereinafter referred to as a technology document collection set) 1) market share, 2) concentration rate, 3) activity rate, and 4) information on other patent analysis indicators. At this time, it is natural that the documents corresponding to the sub-patent classification symbol of the given patent classification symbol in the data processing for the above 1) to 4) are included in the analysis target document set.

In the analysis of patent technology classification, firstly, the analysis of the total amount will be described. The total amount analysis is intended to provide a quantitative analysis result such as the number of applications / registration numbers per period / period by level of the patent technical classification code, and the documents including the sub-patent classification code of the specific patent classification code are rolled up, (The number of patent documents containing a specific patent classification and the number of patent documents containing the sub-patent classification are summed up (it is natural that redundancy is eliminated). Table 58 below is an exemplary implementation of an operation result table for multidimensional analysis on the yearly distribution of the total number of applications for the USPC. USPC 002048000 is 2/48, and 002049100 is 2 / 49.1. This is a matter of the notation of the USPC. (The first three digits are the class number, and the 6 digits after the second digit are divided by 1000.

[Table 58]

ano dot 1 dot 2dot 3 dot 4 dot 01 02 03 04 05 06 002048000 002049100 002049200 4 One 6 3 One 8 002048000 002049100 002049200 4 One 6 3 One 8 002048000 002049100 002049200 4 One 6 3 One 8 002048000 002049100 002049200 4 One 6 3 One 8 002048000 002049100 002049300 0 One One 0 One 0 002048000 002049100 002049300 0 One One 0 One 0 002048000 002049100 002049300 0 One One 0 One 0 002048000 002049100 002049300 0 One One 0 One 0 002048000 002049100 002049400 002049500 0 One 3 One One 2 002048000 002049100 002049400 002049500 0 One 3 One One 2 002048000 002049100 002049400 002049500 0 One 3 One One 2 002048000 002049100 002049400 One 2 6 0 2 4 002048000 002049100 002049400 One 2 6 0 2 4

If you want to know the number of applications / registrations in a specific USPC obtained above, you can obtain the numbers or sums that have been calculated by year / period in the specific USPC level (class to n dot) column.

The multidimensional analysis calculation execution result table generation module 402 may generate the multidimensional analysis calculation result table data as shown in Table 59 below in order to perform analysis according to the applicant's multi-stage patent classification symbol.

[Table 59]

IAppName C3 C4 C5 C6 C7 C8 05 A C12N C12N 15/09 C12N 15/00 C12N 15/10 2 A C12N C12N 15/09 C12N 15/00 C12N 15/11 C12N 15/12 One A C12N C12N 15/09 C12N 15/00 C12N 15/11 C12N 15/31 C12N 15/33 One A C12N C12N 15/09 C12N 15/00 C12N 15/11 2 A C12N C12N 15/09 C12N 15/00 C12N 15/10 2 A C12N C12N 15/09 C12N 15/00 C12N 15/11 C12N 15/12 One A C12N C12N 15/09 C12N 15/00 C12N 15/11 C12N 15/31 One A C12N C12N 15/09 C12N 15/00 C12N 15/11 2 A C12N C12N 15/09 C12N 15/00 C12N 15/10 2 A C12N C12N 15/09 C12N 15/00 C12N 15/11 4 A C12N C12N 15/09 C12N 15/00 6 A C12N C12N 15/00 6 A C12N 6 B

In the case of the data shown in the above Table 59, when extracting multiple applicants based on C12N 15/00, C12N 15/00 (the IPC subgroup is abbreviated as C4 level) The GID 15 that has rolled up the document count information about the subpopulation classification symbol is given as GID 15, so that when the conditional expression "GID = 15 and IPC = C12N 15/00 and IPC level = C4" The number of documents per eligible applicant (number of documents in GID 15) is shown.

When there is the total amount count information as described above, data such as the operation result table for the multidimensional analysis may be generated for each analysis index such as the occupancy rate, the concentration rate, the activity rate, etc. As described above, The following table data of the operation result for the multidimensional analysis is presented. It will be apparent to those skilled in the art that equivalent operations can be performed on other patent indices such as market share.

Multidimensional analysis operation result table for fusion analysis

Hereinafter, multidimensional analysis operation result table data generated by the multidimensional analysis operation result table generation module 402 for fusion analysis will be described.

The multi-dimensional analysis operation result table generation module 402 of the present invention generates operation result table data for multidimensional analysis as follows for the fusion analysis of the present invention. The multidimensional analysis operation result table generation module 402 receives an arbitrary set of patent documents inputted, set or determined, and then, when a plurality of patent classification symbols included in the document set are combined (If there is only IPC, there are two or more IPCs. If there are IPCs and USPCs, the process is performed if there are two or more IPCs. , And generates data as shown in the following Table 72 for the above-mentioned patent classification symbol.

[Table 72]

Article number AppName Main
C1 Main Cn Sub 1
C1 Sub 1
Cm Sub i
C1 Sub i
Ck date Etc One A 2 B

One patent document includes at least one of the types of patent classification symbols such as IPC, USPC, FI, FT and ECLA, and the patent classification symbol includes at least one information. The data as shown in the above Table 72 shows that there can be one or more essential Main classification codes and optional Sub classification codes based on the document number. The patent classification code included in the document lists the own patent classification code included in the document and all the upper patent classification codes of the patent classification code included in the document. Assuming that its own main patent classification symbol corresponds to the IPC 2 dot subclass, the multidimensional analysis operation execution result table generation module 402 refers to the patent classification symbol mast DB 203, The classification symbol is the C6 level, the main patent classification symbol is inputted to the C6 level, the upper right patent classification symbol of the main patent classification symbol of 1 dot level is inputted to the left C5 level, Enter the patent classification code for the level, and enter the process up to the top level in a series. At least one sub IPC included in the patent document is processed in the same manner as in the main IPC. At this time, since other bibliographic data such as the applicant are information available through the document number, these are optional elements.

Korean Patent Application No. 10-2005-0111868 discloses H04B 7/26 and H04B 7/15 as January 2006. (For the sake of convenience, the subclass portion in the patent classification symbol in the main group and below is omitted.

[Table 60]

Article number AppName Main
C1 C2 C3 C4 C5 C6 Sub
C1 C2 C3 C4 C5 C6 Etc 10-2005-0111868 Samsung H H04 H04B 7/00 7/24 7/26 H H04 H04B 7/00 7/14 7/15

If there are two or more Sub IPCs, the same processing is performed in parallel on the right side of the H04B 7/15 related information as shown in Table 61 below. Korean Patent Application No. 10-2006-0012606 discloses a patent classification code H04B 7/04, H04B 7/155, and H04Q 7/30 granted on Jan. 2006, which will be described by way of example. (AppName etc. are omitted for convenience, and H04Q 7/30 is indicated on the line below.)

[Table 61]

Article number Main C1 C2 C3 C4 C5 C6 Sub 1
C1 C2 C3 C4 C5 C6 C7 10-2006-0012606 H H04 H04B 7/00 7/02 7/04 H H04 H04B 7/00 7/14 7/15 7/155 Sub2 C1 C2 C3 C4 C5 C6 C7 H H04 H04Q 7/00 7/20 7/30 ..

Korean Patent Application No. 10-2005-0042032 includes H04B 7/02 and H04B 7/14 as patent classification information.

[Table 62]

Article number AppName Main
C1 C2 C3 C4 C5 C6 Sub
C1 C2 C3 C4 C5 C6 Etc 10-2005-0042032 SK Telecom H H04 H04B 7/00 7/02 H H04 H04B 7/00 7/14

It will be apparent to those skilled in the art that through the multidimensional analysis operation execution result table generation module 402, the same operations as the multidimensional analysis operation result table data can be similarly generated for USPC, FT, and the like.

The multidimensional analysis calculation execution result table generation module 402 generates a plurality of patent classification symbol pair information as shown in Table 63 from the sequence information of the plurality of patent classification symbols for each level. The types of the plurality of patent classification symbols are described in the description of the same type of patent classification related preprocessing module. In Table 63, only one bibliography, such as the filing date, may be newly added to the side (for example, the date of registration). However, in the case of a joint applicant or co-inventor, What should be newly established is as described above. In other words, the same record is generated only by the applicant or the inventor. Based on such information, it is possible to analyze the applicant's convergence and the inventor's convergence.

To explain this model, suppose that there are only three patent documents: Korean Patent Application No. 10-2005-0111868, Korean Patent Application No. 10-2006-0012606, and Korean Patent Application No. 10-2005-0042032. Of course, it is possible to generate the following data for all patent classification combinations (Ai, Bj) based on all document sets by repeatedly processing all the documents obtained. , Bj) will be able to generate the following table. The combination of the patent classification symbols is 22 pairs (6 + 16) in the pair (4 pairs in Korean patent application No. 10-2005-0111868) and 22 pairs (6 + 16) in the Korean patent application No. 10-2006-0012606 IPC, and Sub IPC) pairs, and one pair in Korea Patent Application No. 10-2005-0042032, and 47 pairs in three patent applications. Although it is naturally possible to express both of these pairs, the concept of the present invention will be described in a manner of replacing the description with respect to all 47 by the table as shown in Table 63 below in which a certain portion is omitted for convenience of explanation.

[Table 76]

Patent classification symbol combination (Ai, Bj) Article number Applicant Filing date Apply weighting policy 1 Apply weighting policy 2 (H04B 7/15, H04B 7/26) 10-2005-0111868 Samsung December 4, 2006 One One (H04B 7/15, H04B 7/24) 10-2005-0111868 Samsung December 4, 2006 One One (H04B 7/14, H04B 7/26) 10-2005-0111868 Samsung December 4, 2006 One One (H04B 7/14, H04B 7/24) 10-2005-0111868 Samsung December 4, 2006 One One (H04B 7/155, H04B 7/04) 10-2006-0012606 Samsung December 4, 2006 1/3 0.75 / 2 (H04B 7/155, H04B 7/02) 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 (H04B 7/15, H04B 7/04) 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 (H04B 7/15, H04B 7/02) 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 (H04B 7/14, H04B 7/04) 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 (H04B 7/14, H04B 7/02) 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 (H04B 7/14, H04B 7/02) 10-2005-0042032 SK Telecom December 4, 2006 One One (H04Q 7/00, H04B 7/00) 10-2006-0012606 Samsung 2006.02.09 1/3 0.25 / 1 (H04Q 7/00, H04B) 10-2006-0012606 Samsung 2006.02.09 1/3 0.25 / 1 (H04Q, H04B) 10-2006-0012606 Samsung 2006.02.09 1/3 0.25 / 1

When data is generated in the combination of the above-described patent classification symbols (Ai, Bj), there may be obstacles in sorting, indexing, and / or roll-up operations. Therefore, it is more preferable to conceptually process combinations of patent classification symbols as shown in Table 64 below, and the multi-dimensional analysis operation result table generation module 402 may generate the plural patent classification symbol pair information as follows There will be.

[Table 64]

From Main IPC From Sub IPC Article number Applicant Filing date Apply weighting policy 1 Apply weighting policy 2 H04B 7/15 H04B 7/26 10-2005-0111868 Samsung December 4, 2006 One One H04B 7/15 H04B 7/24 10-2005-0111868 Samsung December 4, 2006 One One H04B 7/14 H04B 7/26 10-2005-0111868 Samsung December 4, 2006 One One H04B 7/14 H04B 7/24 10-2005-0111868 Samsung December 4, 2006 One One H04B 7/155 H04B 7/04 10-2006-0012606 Samsung December 4, 2006 1/3 0.75 / 2 H04B 7/155 H04B 7/02 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 H04B 7/15 H04B 7/04 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 H04B 7/15 H04B 7/02 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 H04B 7/14 H04B 7/04 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 H04B 7/14 H04B 7/02 10-2006-0012606 Samsung 2006.02.09 1/3 0.75 / 2 H04B 7/14 H04B 7/02 10-2005-0042032 SK Telecom December 4, 2006 One One H04Q 7/00 H04B 7/00 10-2006-0012606 Samsung 2006.02.09 1/3 0.25 / 1 H04Q 7/00 H04B 10-2006-0012606 Samsung 2006.02.09 1/3 0.25 / 1 H04Q H04B 10-2006-0012606 Samsung 2006.02.09 1/3 0.25 / 1

For the purpose of statistical processing and analysis, the multidimensional analysis operation execution result table generation module 402 preferably generates the plural patent classification symbol pair information as shown in Table 65 below. (For convenience of description, But it is the same as the sequence of the document numbers in Table 64. The multidimensional analysis operation result table generation module 402 generates the multidimensional analysis operation result table In the data, it is desirable that the bibliographic number of the document number (including the applicant, the inventor, the filing date, the date of registration, and other bibliographic information such as the registration date) M, and a sub-IPC, it is a field that means that it is represented by S. This indicates that the convergence criterion is preferably the axis of Main IPC It is because that preferred.)

[Table 65]

C3 C4 C5 C6 M / S Article number M / S C1 C2 C3 C4 H04B 7/00 7/14 7/15 S H04B 7/15 H04B 7/26 M H H04 H04B 7/00 H04B 7/00 7/14 7/15 S H04B 7/15 H04B 7/24 M H H04 H04B 7/00 H04B 7/00 7/14 S H04B 7/14 H04B 7/26 M H H04 H04B 7/00 H04B 7/00 7/14 S H04B 7/14 H04B 7/24 M H H04 H04B 7/00 H04B 7/00 7/14 7/15 7/155 S H04B 7/155 H04B 7/04 M H H04 H04B 7/00 H04B 7/00 7/14 7/15 7/155 S H04B 7/155 H04B 7/02 M H H04 H04B 7/00 H04B 7/00 7/14 7/15 S H04B 7/15 H04B 7/04 M H H04 H04B 7/00 H04B 7/00 7/14 7/15 S H04B 7/15 H04B 7/02 M H H04 H04B 7/00 H04B 7/00 7/14 S H04B 7/14 H04B 7/04 M H H04 H04B 7/00 H04B 7/00 7/14 S H04B 7/14 H04B 7/02 M H H04 H04B 7/00 H04B 7/00 7/14 S H04B 7/14 H04B 7/02 M H H04 H04B 7/00  H04Q 7/00 S H04Q 7/00 H04B 7/00 M H H04 H04B 7/00 H04Q 7/00 S H04Q 7/00 H04B M H H04 H04B H04Q S H04Q H04B M H H04 H04B

When there are table data of the operation result table for multidimensional analysis including the above-mentioned fusion information targeting all patent documents including two or more patent classification codes, the following information can be extracted / calculated.

First, it is an analysis of non-use fusion of subject information when a patent classification code is obtained. When one patent classification symbol is obtained or given, one can find a ranking of the patent classification that fits well with the patent classification symbol (high frequency of being paired). In the above table, for example, the frequency of fusion at each IPC level, such as H04B 7/155 at C7 (IPC 3 dot level) level and H04B 7/15 at C6 level, is a patent classification code that fuses well with H04B 7/04 (By group, count, and rank commands), it is possible to obtain ranking information of IPCs with high convergence frequency For example, "the C6 level IPC that fits best with H04B 7/04 is H04B 7/15, and the C6 level IPC that fits well with the next level is H04Q 7/30" Of course, the specific document in which the fusion occurred can be specifically confirmed through the document number, but it is also possible to obtain corresponding documents by querying the search engine or the DBMS 201 for the IPC pair in which fusion has occurred The obtained patent classification Patent classification codes that are well-compatible with the IPC sub-class can be generated by analyzing IPC sub-class, main group, 1 dot, 2 dot, ... according to year / period, (Preferably H04B 7/04) from one side (preferably the main IPC) and the level is determined among C1 to Cn in all the records (rows) including the found patent classification symbol (By group, count, and rank commands) for all the patent classification symbols existing in the determined level, and period / date information is included in the operation result table data for the multidimensional analysis , It is possible to generate statistical / analytical data for the period (for example, by year) for the patent classification symbols that fuse well with the obtained patent classification symbol.

Although the fusion analysis has been described with respect to one patent classification symbol, the fusion analysis when a plurality of patent classification symbols of two or more are obtained is also applicable to a plurality of patent classification symbols obtained from the two or more patent classification symbols , And sum it up. On the other hand, according to the policy, when a plurality of patent documents obtained in one document are overlapped, a statistical / analysis information may be generated while permitting duplication, and statistical / analysis information may be generated while eliminating redundancy. . In the latter case, after querying the plurality of patent classification codes obtained, the duplicate removal instruction such as a distinct instruction is performed on the document number included in each of the records resulting from the query, and then, It is possible to generate statistical / analytical information about the patent classification that is well converged by the IPC level. The reason for such a duplication problem is that when convergence information is processed, if a fusion occurs in a sub-patent classification symbol, fusion is naturally occurred between the respective upper classifications. On the other hand, the plurality of patent classification symbols do not have to be the same IPC level, and the inventive ideology when the plurality of patent classification symbols are obtained can be applied to any two or more patent classification symbols selected from the entire IPC classification system Of course.

Second, analysis of the use of subject information when the patent classification code is obtained. When there is a set of convergence occurrence documents related to the set of patent classification codes that best fuse with the obtained patent classification code (one or a plurality of), the bibliographic information related to the document number of the convergence occurrence document set is used, Field-specific statistical / analytical information of the field. For example, a set of patent documents of Samsung Electronics Co., Ltd., filed with the Korean Intellectual Property Office, which includes a patent classification code that has been merged with H04B 7/00, is set as a document to be analyzed, and the IPCs in the document set are classified Class, main group, 1 dot, 2 dot ... (at this time, the idea of including the subpopulation classification is of course applicable). Documents where certain IPCs are merged by year / The information can be provided for each inventor of Samsung Electronics Co., Ltd.

Third, it is a fusion analysis within a given set of documents. The set of documents given above may be generated by a combination of at least one of the various fields constituting the bibliography such as 1) applicant, 2) inventor, 3) patent classification symbol, 4) country, 5) date. When a set of documents is obtained, the set of documents can be identified from the obtained document set by querying the operation result table for multidimensional analysis including the fusion information. (A record including the document number and the document number is specified). Here, the information that can be generated for the document set in which the fusion occurs is as follows. 1) It is possible to extract IPCs for each IPC level from the document set in which fusion occurs, and generate statistical / analysis information for IPCs for each level in which the convergence occurred, by the extracted optimal IPCs. The information is counting information. If the counting information value is present, the rate of change, rate of increase, and the like can also be obtained.) 2) Statistics / analysis data on the first half of the present invention, (The ranking information of the most inventors, the most applicants, the most convergent patent classification symbols, etc.)

Fourth, statistics / analysis for fusion target discovery. This means that a criterion is found within a given set of documents and that a convergence of all or a predetermined set of documents of the first or second country (for example, a set of documents filed during the last seven years on the filing date) This is statistical / analytical analysis that obtains convergence statistics / analysis information for each criterion by querying the operation result table data including information. For example, the mode of the patent classification symbol level by total applicant or applicant total amount / market share / concentration ratio / activity rate or other analysis index. More specifically, it is possible to extract, for example, main group standard consensus IPCs of Samsung Electronics Co., Ltd. in Korea, and to extract reference IPCs . At this time, the operation result table data for multidimensional analysis including the fusion information, which is the object to be queried, may be the first station (the country in which the criterion is generated), but the second station ) Was created from a set of patent documents. Using the above method, the IPC for each main group is extracted from Samsung Electronics' patent documents of Korea, and the multi-dimensional analysis including the convergence information generated by the patent document set of Japan or USA is performed through the extracted IPC It is possible to obtain the patent classification symbol information for each stage that is well-integrated with the IPCs of the concentrated main group, which is extracted in Japan or the United States, where the fusion frequency is high, by querying the result table data.

In the present invention, generation of operation result table data for multidimensional analysis by roll-up processing mediated by a sub-patent classification symbol on a multi-stage patent classification system is described. Also, even if the roll-up is presupposed or even if it is not presupposed, information about the sub-classifications of a given patent classification symbol is generated in real time, and information about the sub- To the user. The inventive idea of the present invention is applied to fusion analysis as it is. The following provides a more detailed description of drill down in the fusion analysis.

The reference IPC in Table 79 below is the IPC obtained or given, and the target IPC is the set of convergent documents extracted from either the first or second total convergence document set or the previously established or obtained document set, It is the most frequent / frequent IPC obtained from the convergence document set. At this time, it is of course possible to select the level of the reference and target IPCs. (Of course, the reference IPC is a set of convergence documents extracted from either a first or second country full convergence document set or a pre-established or acquired document set, or a set of pre- If there is an independent IPC level selection from the user for each of the reference IPC and the target IPC, the selection information is obtained, and the calculated result for each selected level is extracted and provided to the user .

[Table 66]

ranking Standard IPC
Target IPC IPC Description 00 01 lately Sum One B60C 3/00 (+)
H01B 1/00 (+) 11 52 4 2 H03M 3/00 (+)
B29C 31/00 (+) 2 5 9

When the reference IPC B60C 3/00 in Table 66 is pressed to obtain drilled-down convergence statistical / analysis information, the system 1 of the present invention generates drilled-down data in the manner shown in Table 67 below You will be able to provide it to users. (Drilldown is only available to subordinates under the IPC scheme, and not all of the IPCs that are available appear, but only IPCs that have converged.)

[Table 67]

ranking Standard IPC
Target IPC IPC Description 00 01 lately Sum One B60C 3/00 (+)
H01B 1/00 (+) Tires characterized by cross-section

A conductor or a conductive object characterized by a conductive material 11 52 4 67 B60C 3/04 (+) A feature of the cross section method of section 5 30 3 40 B60C 3/06 (+) Asymmetric 4 10 One 12 B60C 3/08 (+) Foldable in storage or unused condition One 2 0 2 2 H03M 3/00 (+)
B29C 31/00 (+) 2 5 9 16

(Remarks) 52 = from child (30 + 10 + 2) + self (10)

If the user of the present system presses H01B 1/00 anywhere in the result data as shown in Table 67, the system 1 can generate drilled-down data of the format shown in the following Table 68 and provide it to the user.

[Table 68]

ranking Standard IPC
Target IPC IPC Description 00 01 lately Sum One B60C 3/00 (+)
H01B 1/00 (+)
Tires characterized by cross-section 11 52 4 H01B 1/02 (+) Mainly of metals or alloys 3 35 2 H01B 1/04 (+) Mainly made of carbon silicon, carbon or silicon 2 10 0 H01B 1/06 (+) Mainly made of other non-metallic materials 2 0 0 2 H03M 3/00 (+)
B29C 31/00 (+) 2 5 9

On the other hand, the system 1 can provide the following result data to the user in relation to the drill-down. The reference IPC in Table 69 below is the IPC obtained or given, and the target IPC, which is ranked from the first to the nth ranked, is extracted from the aggregation document set of the first or second whole convergence document, The most frequent / frequent IPCs that are obtained from a set of established convergence documents or from a set of pre-established or acquired convergence documents. (Of course, the reference IPC is a set of the convergence document collection of the first or second convergence document collection or the document collection set of the pre-established or acquired document set, Or it may be the most frequent / frequent IPC available from a set of pre-established or acquired convergence documents.)

[Table 69]

ranking Standard IPC IPC Description 1st 2nd place 3rd place One B60C 3/00 (+) Tires characterized by cross-section H01B 1/00 (67) (*) G03B 5/00 (15) (*) C07B 31/00 (10) (*) 2 H03M 3/00 (+)

At this time, when H01B 1/00 (*) is pressed in the result data as shown in Table 69, it is preferable that the result of the fusion analysis of the reference IPC and the target IPC as shown in the following Table 70 is displayed.

[Table 70]

ranking Standard IPC
Target IPC IPC Description 00 01 lately Sum One B60C 3/00 (+)
H01B 1/00 (+) Tires characterized by cross-section
Tires characterized by cross-section (more)

A conductor or a conductive object characterized by a conductive material

11 52 4 67 H01B 1/02 (+) . Mainly of metals or alloys 3 35 2 40 H01B 1/04 (+) Mainly made of carbon silicon, carbon or silicon 2 10 0 12 H01B 1/06 (+) . Mainly made of other non-metallic materials 2 0 0 2

It would be desirable for all IPCs visible to users to include an IPC description. It is difficult to expect to store all of the IPCs among the users, so that it is preferable that the title information of each IPC firstly appears in the IPC description information. At this time, since title information of each IPC is brief and explains only the characteristics, when the description of the IPC is clicked in the second place, the upper IPCs of the IPC are all up to a predetermined level (for example, Level).

The method for generating multidimensional analysis operation result table data for the fusion analysis based on the IPC has been described in the USC, FT, FI, ECLA, etc. It will be apparent to those skilled in the art that other patent classification schemes of the invention may be applied in an equivalent manner.

In the fusion analysis, the combination of the patent classification codes generated by fusion is described based on the combination of two patent classification codes (Ai, Bj). However, in the present invention, combinations of three or more fusion patent classification codes (Ai, Bj, Ck ...) will be obvious to those skilled in the art. For example, in the case of fusion analysis based on the combination of the two patent classification symbols, a patent document including two or more similar patent classification symbols is defined as a set of fusion patent documents. However, , A patent document containing three or more patents of the same kind should be used as a set of fusion patent documents. In the case of a combination of two patent classification codes, there are nC2 (Ai, Bj) combinations when there are n patent classification codes in one document. However, if there are three patent classification codes, , Bj, Ck), and (Ai, Bj, Ck) can be processed as one unit like (Ai, Bj). In this case also, any one of Ai, Bj, and Ck may be originated from the main IPC, or all may originate from the sub IPC.

Analysis result table for multidimensional analysis for representative phrase analysis

A representative word generation method using the representative word extraction preprocessing module of the present invention has been described above. The representative word extraction preprocessing module generates data as shown in Table 71 below.

[Table 71]

Phrase ID Fish Article number field Number of times of appearance One abc #One D 3 One abc #One C 2 One abc #2 D 2 One abc #2 C One 2 bcd #One D One 2 bcd #One C One 3 ac #One D One

The multidimensional analysis calculation execution result table generation module 402 of the present invention acquires data as shown in Table 84 to generate multidimensional analysis calculation result table data for analysis of representative phrases, (Patent applicant, patent classification code, inventor, date information, etc.) included in the bibliography of the patent document master DB 202 (the patent document master DB 202 may perform processing by performing a join operation with the document number as a key value). The multidimensional analysis operation result table data of the form shown in Table 72 below is generated. First, the patent classification code can be extracted to generate the following data. For convenience, suppose document # 1 contains one IPC, H04B 7/15, and document number # 2, contains IPCs H04B 7/26 and H04B 7/02.

[Table 72]

Phrase ID Fish field Number of times of appearance Article number C1 C2 C3 C4 C5 C6 One abc D 3 #One H H04 H04B 7/00 7/14 7/15 One abc C 2 #One H H04 H04B 7/00 7/14 7/15 One abc D 2 #2 H H04 H04B 7/00 7/24 7/26 One abc C One #2 H H04 H04B 7/00 7/24 7/26 One abc D 2 #2 H H04 H04B 7/00 7/02 One abc C One #2 H H04 H04B 7/00 7/02 2 bcd D One #One H H04 H04B 7/00 7/14 7/15 2 bcd C One #One H H04 H04B 7/00 7/14 7/15 3 ac D One #One H H04 H04B 7/00 7/14 7/15

Since document number # 2 corresponds to IPC H04B 7/26 and H04B 7/02, a record for processing two pieces of IPC information is generated for each record. Although the IPC has been described above, it will be apparent to those skilled in the art that the present invention is applied to USPC, FT, etc. in the same manner as described above.

First, the patent classification symbol can be extracted and the data as shown in Table 73 below can be generated. For convenience, suppose that applicant A is included in document number # 1 and applicant B and C are included in document number # 2. It is also assumed that the document No. 1 is filed in May 2005, and the filing filed in January 2006. In this case, the multidimensional analysis operation result table generation module 402 of the present invention can generate multidimensional analysis operation result table data of the form shown in Table 73 below. (05 / 1Q in the following table refers to the first quarter of 2005)

[Table 73]

Phrase ID Fish field Number of times of appearance Article number 05 05 / Q1 5 / 2Q 5 / 3Q 5 / 4Q 06 06 / 1Q AppName One abc D 3 #One One One A One abc C 2 #One One One A One abc D 2 #2 One One B One abc C One #2 One One B One abc D 2 #2 One One C One abc C One #2 One One C 2 bcd D One #One One One A 2 bcd C One #One One One A 3 ac D One #One One One A

Since the applicant B and C are included in the document number # 2, there is a record for processing two pieces of applicant information for each record.

The multidimensional analysis operation result table generation module 402 of the present invention generates the two multidimensional analysis operation result table data related to the representative word as one integrated multidimensional analysis operation result table data It might be. On the other hand, if a translation is applied, an operation result table data for multidimensional analysis using the translations of each language corresponding to the language information for the phrases can be generated. The AppName field indicates an applicant's name, an inventor field can be generated next to the field, and one record can be generated for each individual inventor. If the applicant is 2 and the inventor is 3, then one record without applicant and inventor information is extended to 6 (2 * 3 = 6) records containing one applicant and one inventor.

If there is data such as those in Tables 72 and 73, the following information can be generated with respect to the representative word.

3) a specific period (year, quarter, month, etc.); 4) a specific country or territory of a specific country (eg, It would be possible to provide a set of phrases. For example, a list of phrases appearing in a document filed by Samsung Electronics Co., Ltd. (in the claims or abstracts or in a specific field of a document) is extracted from IPC Main Group H04B 1/00, 2005, You can do it. It will be apparent to those skilled in the art that when there are table data of operation result data for multidimensional analysis as shown in Table 72 and / or Table 73, it can be processed by a simple SQL statement.

2) a specific level of a specific patent classification (eg IPC main group); 3) a specific period (year, quarter, month, etc.); 4) It is possible to extract and provide a representative phrase set. The representative phrase extraction depends on the representative phrase extraction policy. Examples of the representative phrase extraction policy may include the following.

1) For a specific phrase extracted, "Occurrence frequency probability = Occurrence frequency in a set of specific documents / Occurrence frequency in the set of its own document * 100" is included in the preset first narrow range value .

2) a value obtained by processing the appearance frequency probability of the specific phrase for a specific phrase is not included in the predetermined first narrow range value but is included in the predetermined second wide value, And is included within the set rankings. This can be applied to cases where the number of representative phrases is remarkably small as a result of using the representative phrase extraction policy of the above 1) in a specific document set.

3) For the extracted specific phrase, the increase / decrease rate of the occurrence frequency in the specific document set is included in the first narrow range value set in advance.

When finding the probability of occurrence, the upper set of documents is referred to. The types of the upper set of documents are as follows. 1) When an element forming the document set includes at least one or more dimensions, (Of course, it is possible to have an upper level or all upper level of two or more levels at a time), 2) when a document set is specified only by a pure keyword, the entire document set There is a set of superscripts that is a set (in the case of a set of national documents, a set of all documents of that country). The upper document set of the above 1) type will be described in detail. For example, in the case of a set of documents characterized by IPC H01L 1/00, a set of documents identified by i) H01L or their top patent classification codes (H01, H) may be the set of documents above. The top set of documents specified by IPC H01L 1/00 of Samsung Electronics Korea Co., Ltd. in the first quarter of 2000 includes i) the top patent classification symbol, ii) the year 2000 or year integration, iii) all countries of Samsung Electronics Co., Translation of the above phrase into the corresponding language may be presumed), iv) all applications from the Republic of Korea, or v) any combination of any of i) to iv) A set of documents generated by the user can be the above document set. When the document set is generated by a combination of a specific keyword and the dimension (for example, the keyword "RFID tag" is combined with the country, applicant, year dimension such as year 2000 of Samsung Electronics Co., Ltd. of Korea) Of course, if i) to v) above are applied, it is of course possible to generate a super document set. On the other hand, it is preferable that the first narrow range value and / or the second wide range values are determined for each set of upper documents of i) to v).

On the other hand, when extracting a representative phrase generated from a document set generated by including at least one dimension in a specific element of a document set, it will be possible to drill down on the dimension as an axis. When drilling down, it is possible to extract a representative phrase that satisfies predetermined conditions for each dimension element and provide the result to the users. Tables 74 and 75 below illustrate one embodiment of a drill down involving representative phrases. The above example shows representative phrases for 00 years. At this time, if you click 01 year, the representative phrase corresponding to 01 year is extracted and shown as Table 74.

[Table 74]

ranking Standard IPC
Target IPC IPC Description 00 01 02 03 04 05 06 lately Sum One B60C 3/00 (+) Tires characterized by cross-section Slip, Improved Friction, Tire Resonance Sound Frequency, Noise, Static Emission
2 H03M 3/00 (+) Modulation, sigma delta, pulse width modulation, red-eye reset

It is preferable that the representative word is different in color, font size and the like according to the appearance frequency probability. On the other hand, when the representative phrase is clicked on, the document containing the representative phrase can be extracted. The document extraction is performed by: 1) extracting the corresponding document by querying the patent document master DB 202 based on the document number if the document number is present in the multidimensional analysis operation result table data for extracting the representative word, , And 2) there is a method of querying a search engine by combining all dimensions and representative phrases from the above. In the latter example, if the representative phrase "noise" is selected (if this representative phrase is from the claim (C), the search formula entered is "IPC = B60C 3/00 AND Claim = , Date = (20000101 ~ 20001231 ", Country = Republic of Korea".) If you want to see US documents with this representative phrase, you can query the search engine by replacing "Country = US" If the "noise" corresponding to the keyword is not English, which is the basic language of US documents (ie, judging the type of language), "noise" is translated into English using the translation system included in this system (1) it is of course possible to input and search for the "noise"

The system 1 of the present invention may be configured as shown in Table 75 as below, in order to extract the drilled-down representative phrases, if the reference IPC B60C 3/00 is pressed in Table 74 The data can be generated and provided to the user. (Drill down is only available to subordinates under the IPC scheme, not to all of the IPCs that are available.) For example, if there is no document in B60C 3/02, or there is no representative phrase, the representative for this IPC At this time, the representative phrases corresponding to B 60 C 3/00 may be shown or hidden.

[Table 75]

ranking Standard IPC
Target IPC IPC Description 00 01 lately Sum One B60C 3/00 (+) Tires characterized by cross-section Slip, Improved Friction, Tire Resonance Sound Frequency, Noise, Static Emission B60C 3/04 (+) A feature of the cross section method of section Adjustment stability, radial tires, compressed air B60C 3/06 (+) Asymmetric Abrasion resistance, cornering, asymmetric self-sidewall, asymmetric radial tire B60C 3/08 (+) Foldable in storage or unused condition Rim release, fair tire carrier, storage device 2 H03M 3/00 (+) Differential modulation

It will be appreciated that the above-described example is merely an example of presenting representative phrases, and the representative phrases can be extracted and provided at any place of the analysis result to which the present invention is applied. This is because the analysis result to which the idea of the present invention is applied corresponds to the document set, and the representative phrase of the present invention is generated by obtaining the document set (the patent documents constituting the document set).

The multidimensional analysis operation result table generation module 402 of the present invention generates a multidimensional (n-dimensional) cube by performing a multidimensional operation calculation for each selected at least one dimension. One of the features of the multidimensional cube generation of the present invention is that when the cube is generated by including the patent classification symbol in the dimension, when the roll-up operation value for the given patent classification symbol is generated, the patent classification symbol and the patent classification symbol And generate the rollup operation value by considering the upper patent classification symbol. When the roll-up operation value is generated, when the roll-up operation value for an arbitrary patent classification symbol is obtained, all values of the sub-patent classification symbols of the patent classification symbol are reflected by the multidimensional numerical values, . The multidimensional analysis operation result table generation module 402 of the present invention generates the multidimensional analysis operation result table generation module 402 in accordance with the patent classification notation Reflects the processing value for the symbol. For example, when an IPC named H04B 7/06 is assigned to document number # 1, a count value of 1 is assigned to H04B 7/06 when generating an operation result table for cubic multidimensional analysis using this document. H04B 7/06 should also be given a count value in the upper hierarchy H04B 7/04, H04B 7/02. Of course, it is common sense that counting values are also given to higher levels of H04B 7/00 and higher. For this purpose, as in the above tables, for each given patent classification code (for example, IPC), the upper-level patent classification codes of their own are arranged in a row on the basis of their own patent classification codes, The cell value is calculated by reflecting the value derived from the given patent classification code. The unique roll-up operation method according to the present invention is repeatedly described in the operation result table generation module 402 for multi-dimensional analysis and throughout the description.

Self-set criterion Self-set analysis and self-criterion Other type of analysis

The analysis module of the present invention is composed of a set of various analysis subject-specific analysis modules. The analysis subject-specific analysis module includes an analysis subject-specific query expression corresponding to the analysis subject, in order to generate an analysis result related to the subject. Dimensional analytical operation result table generated by the multidimensional analytical operation result table generation module 402 in the analytic subject-specific analytic query expression, and obtains the subject analysis results.

The analysis result value generated by the analysis module of the present invention and the initially determined document to be analyzed have the following two relationships.

The first is the case where the result of the analysis is derived from the document set to be analyzed as the "self-set reference self-set analysis type ". For example, the number of applications, market share, concentration rate and activity rate data for each stage of IPC of a specific applicant, and the number of applications, market share, and concentration of each applicant by IPC Data on activity rate data, data on the number of applicants, or the number of persons cited by individual inventors (technical fields), and the like are examples thereof. At this time, the most specific technical field of the specific applicant is extracted. The IPC is extracted from a set of documents related to the specific applicant himself / herself, and the extracted IPC is used as a standard, The analysis results such as the number of applications and market share by year are created. Therefore, it is created based on the set of documents of one's own, and the final analysis result of each criterion is the number of applications per year, so it becomes an analysis of oneself. The above-mentioned "self-set criterion self-set analysis" type is characterized in that the output result of the final analysis can be generated from the first set of document sets (self-document set). (Of course, All values are required.)

The second case is a type of "self-set criterion other-set analysis", in which the above-described analysis result value does not appear in the analysis target set. For example, if you want to obtain data about the citations of applicant A by multiple application technology fields, the first data you input is related to the document set of applicant A, but the final output data Is data about the document set of the non-A person. Also, in the case where the applicant who has been filed in the IPC or the applicant who has a high concentration rate in the IPC is extracted by the IPC of his or her most recent main group stage, the criteria of "self- The IPC of the main group is extracted from the document set of the main group and the final result is the set of documents of the others.

In this second case, two or more calculations are required. In the above example, first, a set of documents of applicant A is outputted for each of the technical fields of application, and a set of documents of the other persons quoting documents belonging to the set of documents for the output set of documents for each technical field is obtained, A set of documents is extracted as a set of documents to be analyzed, and the applicants are extracted as final analysis results.

The multidimensional analysis operation result table generation module 402 may generate the same multidimensional analysis operation result table from the multidimensional analysis operation result table to the predetermined final result . For example, one multidimensional analytical operation result table is often sufficient to generate analytical results such as number of applicants or market share per applicant, by IPC per year.

On the other hand, in the case of the self-set reference other-set analysis type, 1) when the final result is generated up to the final result, 2) the result up to the final result is divided into each step, There can be a way. In the latter case, the multi-dimensional analysis operation execution result table generation module 402 (hereinafter, referred to as &quot; multi-dimensional analysis operation result table generation module &quot; ) Has multivalued analysis operation result table data or cube data for calculating an output value for the input value, and 3) the multidimensional analysis operation result table generation module 402 inputs the first and each step result The same conditions must be met as you can get by value.

Hereinafter, an example will be described in which the applicant A obtains data on an applicant (applicant, assignee) citing a plurality of documents by application field. The above results can be obtained through the following steps. 1) the stage where Applicant A is confirmed (since Applicant A is one of the entire set of all applicants, A is confirmed); 2) the stage in which the document set of A in one country (eg, the United States) (Eg, IPC, subclass level), because all the document sets of A have at least one IPC or USPC, and so on. (For example, IPC H04N, H01B, etc.) belonging to each class and level are established, and if not given, given the given value, (B) extracting the rankings of the multi-application technology fields by counting the subset of documents for each technical field; (5) for each subset of documents (eg for IPC H04N of all documents of A) (6) extracting other document numbers citing the document number for each individual document, (7) extracting the document number of the extracted other document number (8) collecting the applicants of the extracted other document numbers by the partial document set (the applicant's frequency values are stored together), 9) sorting the combined applicants by the technical field (And / or other document number information) for each applicant is also stored), the applicant A's documents are classified according to the multi-filed technical field (patent technical classification and level) A number of applicants may be identified. The multidimensional analysis operation result table generation module 402 may perform the steps 1) to 9) with respect to any one or more applicants selected from among all applicants, and then store the resultant values as data . 4) to 5) correspond to the self-analysis of the first self-entry criterion, and 6) to the first self- 7) to 9) relate to the generation of final analysis results corresponding to the final output. In this case, the analysis module may fetch the final analysis result generated by the multidimensional analysis calculation execution result table generation module 402 according to the obtained conditions. The condition may be the type and level of the patent technical classification such as the IPC inputted from the user or the like, the country, and the like.

When the steps 1) to 9) are performed through the multistage analysis, the following grouped multistage method will be possible. In this case, the analysis module may be operated in the following manner.

First, the analysis module establishes a document set for each application field of applicant A. At this time, in the analysis module, the multidimensional analysis calculation execution result table generation module 402 performs the processes of 1) to 4) for all applicants to determine ranking of the multi-application technology field and a subset of documents Data on applicant A can be obtained from the number-confirmed data.

Second, the document collection acquisition module of the analysis module obtains the document number of 5). When there is a number of the document subset, all the numbers can correspond to the SQL query, so that the document number of the 5) can be determined by the SQL query. The document number may be stored in advance in the multidimensional analysis operation execution result table or the cube data, but it would be reasonable not to store it (because it can be called up by SQL at any time). The document collection module receives a document set according to predetermined conditions and performs a union operation. In this case, SQL is the condition for obtaining the document set.

Third, the document collecting module of the analysis module performs step 6). The document collection module receives the document number of step 5) or queries the search engine to obtain the document collection of step 6). Of course, in the bibliographic database, each document includes information (reference information) about the other document number that the user himself / herself is referring to, or when a specific document number is inquired to the search index 401-2, (That is, when indexing is performed, a document has # 1 and # 2 document numbers as citation information), the document number should be able to be output as the reference number. When the search indexer 401-3 generates the inverse file information such as a <- # 1 and a <- # 2 for the search of the quoted field and processes it as the search index 401-2, If you enter the document number # 1 in the field, it prints a number of documents including the document a that quotes itself.)

Fourth, a set of other documents to be analyzed is generated through the step 6), and various analyzes after the generated document set 7) are performed by the direct analysis module of the present invention. The analysis module calls the direct analysis module included in the analysis module to perform steps 7) to 9). The direct analysis module refers to an analysis engine that performs a predetermined analysis process when given a set of documents for a given set of documents (ie, a set of documents that can not be pre-set, It is impossible or unreasonable to preliminarily generate the analysis result of the document set in the case where the number is too large. Therefore, it is reasonable that the multidimensional analysis operation execution result table generation module 402 generates only the step 4). ) The role performed by the direct analysis module is equivalent to the role performed by the operation result table generation module 402 for the multidimensional analysis. That is, through the step 5)

The direct analysis module and the operation result table generation module 402 for multidimensional analysis are different from each other only in the initial stage in which the document set to be analyzed is input, and the analysis result data generation process performed after the document set is once acquired is the same .

Types of self-set criteria Other sets of analysis

The above-mentioned self-set reference other-set analysis will be described in more detail. The first step of the above-mentioned self-set reference other-set analysis is to extract a reference to a self-set. Based on the self-document set, the above criteria include 1) a patent classification symbol for each level with many applications or registrations, 2) a patent classification symbol for each concentration level / occupancy rate / activity rate, and 3) A patent classification code for each level, 4) a patent classification code for each level, 5) a patent classification code for each level in which a lot of convergence occurs, and 6) an average number of claims, etc. . The above 1) to 6) may be extracted for each predetermined period (for example, the last 7 years based on the filing date), and 1) to 6) (1) to (6) above may also be extracted on a nation-by-country basis, provided that the variation value (increase / decrease rate, change rate, etc.) of the family size, It will be possible.

The ranking first place H01L, the second place H04N, etc. shown in Fig. 17 and the like show an example of the criterion extracted for the self-assembly. The self-set refers to a set of documents determined according to the determination of the document set described in this specification.

The second step is a step of generating an analysis result for a criterion extracted for the self-set. In order to generate the analysis result, the analysis subject-specific analysis module queries the operation result table for multidimensional analysis predetermined by a predetermined analysis quality consciousness. At this time, the operation result table for the multidimensional analysis is generated for a set other than the self-set or a whole other set that may include a self-set. . For example, in the case of a set that is not a self-set, 1) When the self-set is a set of the first country, the document set of the second country is representative. On the other hand, the other whole sets, which may include the above-mentioned self-aggregation, include (i) a set of entire frontal citation documents used in citation analysis, (ii) a set of frontal citation documents, A collection of documents is representative.

On the other hand, there is competition analysis as a representative example of the other set analysis of the self-collection standard. Describe competition analysis by taking applicant unit competition analysis as an example. A plurality of patent applications of the applicant in at least one or more patent classification level units (for example, an IPC subclass) with the document set as a self-assembly, targeting a set of documents of all or a part of specific applicants of the first country, It is possible to extract IPCs with high ranking according to multi-registration standard IPC, high concentration ratio, high occupancy rate, high activity rate or other patent indicators. It is possible to extract the rankings of the above-mentioned multi-application / multi-registration applicant, the applicant having a high concentration rate, the applicant having a high occupancy rate, the applicant having a high activity rate or the applicant having a high patent index in the first country or the second country with respect to the extracted IPC There will be. On the other hand, even if it is not the applicant unit, it is possible to carry out the competition analysis for each unit of the applicant's unit, the inventor's unit, and the representative unit by the same processing of the information carried out by the applicant unit in this paragraph. Of course, in order to generate the above-mentioned other set analysis result, the analysis module of the present invention includes a plurality of IPC calculation levels for each level of the IPC, including a total of application / registration amount, occupancy rate, concentration rate, The analysis results are retrieved by approaching the set analytical quality consciousness.

The analysis module queries the operation result table for the corresponding multidimensional analysis in the analytical subject-based analysis query expression, obtains the analysis result, and outputs the analysis result. At this time, the output of the analysis result is preferably output for each language.

On the other hand, when a numerical value is given, the change values (increase / decrease rate, acceleration, acceleration) and statistical values (average and standard deviation, etc.) of the numerical value can also be useful information. In this case, 1) a method of generating the change value and the statistic value in the operation result table for multidimensional analysis may be applied, and 2) there may be a method of instantly calculating the numerical value obtained from the analysis module. At this time, if the numerical value is a counting value such as a simple application / registration amount, the method 2) may be easy, and if the ratio value such as the occupancy rate, the concentration rate and the activity rate, will be. For example, the total number of applications, the number of applications, and the rate of increase / decrease in the number of applications from 2000 to 2005 in the technology field of Company A H04N 7/02 can be easily calculated by the number of applications per year since the numerical values are counting information. On the other hand, when the average value (the sum of the ratio values is meaningless) is calculated by the numerical value of occupancy rate, concentration rate, and activity rate in year H04N 7/02 of company A from 2000 to 2005, Since the sum of the values does not make sense, the formula corresponding to each analytical index should be applied integrally throughout the year from 2000 to 2005, not by year. Therefore, it is preferable that the ratio value is generated at the time of generation of the operation result table for the multidimensional analysis in advance as in the method 2).

The analysis module of the present invention includes a mast DB analysis module that targets the DB itself such as a patent document mast DB 202, a patent classification symbol mast DB 203, and a subject mast DB 204. The mast DB analysis module is provided for each topic. The mast DB analysis module performs an SQL query statement containing information on how to process a value obtained by joining a table of a mast DB for each analysis subject, It contains an application that reprocesses data obtained by SQL query statements. The mast DB analysis module should use a plurality of subqueries. The subquery is a part of the multidimensional analysis operation result table equivalent to the one generated by the multidimensional analysis operation result table generation module 402 One record or a plurality of records) or the entire content.

The multidimensional analysis operation execution result table generation module 402 and the sub-query will be described in more detail by taking as an example a case where a set of user-created documents generated by a user inputting a search word or the like is set as an analysis target document set . Since the document set is not determined in advance before the user creates the document set, the multidimensional analysis operation result table generation module 402 can not generate the multidimensional analysis operation result table in advance there is a problem. Therefore, when analyzing a set of user-created documents, it is necessary to use one or more of the following two methods after confirming the document set to be analyzed. First, the multidimensional analysis calculation execution result table generation module 402 generates a multidimensional analysis calculation execution result table on the user-generated document collection set, and performs various analysis The subject-specific analysis module accesses the table with the corresponding analysis quality consciousness, and generates the analysis result. Second, the mast DB analysis module generates data (table or view) equivalent to the data included in the operation result table for multidimensional analysis using at least one or more sub-queries for documents included in the user-created document set And a mast DB analysis module according to the analysis subject matter accesses the data generated by the analysis query consciousness (main query) to generate the analysis result. The data equivalent to the data included in the operation result table for multidimensional analysis refers to data equivalent to a result of performing a difference operation on at least one dimension such as a roll-up operation or a cube operation.

Direction of analysis series

The multi-dimensional analysis operation result table generated by the multi-dimensional analysis operation result table generation module 402 may have the following sequence. Analysis is basically to obtain the desired data from specific data, so the type of analysis is determined by 1) the attribute of the document set to be analyzed, 2) the subject of analysis, and 3) the form of outputting the analysis result It is decided according to. Therefore, naming is performed according to the types of analysis according to the above 1), 2) and / or 3).

Hereinafter, the various kinds of analysis will be exemplarily described and related to the above 1), 2) and 3).

First, the attributes of the document set to be analyzed are determined according to the criteria of the determination when a set of documents mentioned in this specification has been identified. The applicant, the inventor, the patent classification code of a certain level, the set of forward quotation documents, etc. all depend on the attribute of the document set to be analyzed. These sets of analysis target documents are an example of a unit of a document set obtained by the multidimensional analysis calculation execution result table generation module 402 of the present invention.

According to the analysis subject, the multidimensional analysis calculation execution result table generation module 402 determines which multidimensional analysis calculation execution result table is to be generated and the multidimensional analysis calculation execution result table 402, Which will determine what kind of analytical ques- tion will be asked.

Third, there are two kinds of classification according to the type that outputs the analysis result. 1) single analysis and 2) comparative analysis. In a single analysis, one kind of analysis result is presented as one single criterion, and comparative analysis is a kind of (same kind) analysis result is presented as two or more criteria. For example, a comparison of applicants / companies by two or more applicants / companies (eg, Samsung Electronics and LG Electronics) showing the same analysis result (for example, the number of applications in certain countries by IPC level by year) It is the simplest example of analysis. In this case, it can be seen that the number of applications in a specific country by IPC level by year is the same type of analysis result for each of the two or more applicants / companies. Since the result of the final output to the user is changed according to the selection of the comparison object, the selection result of the comparison object is obtained from the user, and the analysis subject matter analysis You can query by quality consciousness, and then combine the analysis results for each selected item. In the above example, if the same analysis results are generated for Samsung Electronics and LG Electronics respectively and are combined in a manner of comparing them, a result of the comparative analysis is generated.

The monitoring module (403)

Monitoring of patent information refers to monitoring, monitoring, and alerting activities on updated patent information.

To perform the monitoring service, 1) registering a search formula for monitoring; (For example, searching for a new document on the basis of a date including the date of inquiry on the basis of a public date), the query is searched in the search engine or the DBMS 201 The work is limited to the time range) is limited to the search formula), and if there is a new patent document, it may be transmitted to the surveillance service recipient. A special form of surveillance service can be constructed by constructing a tree structure of the classification system of technology, enterprise, product, etc. shown in Fig. ***, and combining a keyword search expression with each node constituting the tree .

The monitoring search formula input to the monitoring module 403 of the present invention is pre-set in the unit patent information service system of the applicant and / or the unit patent information service system 1 of the inventor. That is, when the unit patent information service system of the applicant's name or the unit patent information service system of the inventor is created, or after the generation thereof, the necessary surveillance search expression or the unit of the inventor's name in units of the applicant's name The search formula is automatically extracted and set in the monitoring service of the unit patent information service system of the applicant's name or the unit patent information service system of the inventor's name. The type of the monitoring search formula to be set and the method of generating the monitor search formula may be as follows. First, the type and generation method of the monitoring search formula set in advance in the document set unit will be described. The information to be included in the monitoring search formula may be the analysis result generated by the analysis module of the present invention. Examples of such values include a mode value such as the numerical value and the rate value, a maximum value, and the like. Specific examples are described one by one.

Unit monitoring of applicant's name or inventor's name

First, at least one or more of the most preferred patent classification symbols are ranked in order of frequency from the document set to be managed by the unit patent information service system of the applicant's name or the unit patent information service system of the inventor's name, And may be registered as a search formula to be monitored by at least one country selected from the set country or the country where the document acquired by the patent information mast DB (including WIPO).

1) at least one patent application filed or registered at home and abroad, 2) a patent application filed on or more than a certain number of domestic patents, 3) The applicant whose patent application rate or patent registration rate is higher than the predetermined standard within the predetermined period set in domestic and foreign countries can be extracted and the applicant can be made into the monitoring search formula.

A patent classification code having a high rate of increase in the number of new documents within a predetermined recent period corresponding to the patent classification code and the subpublic classification code of the patent classification code for at least one or more of the most preferred patent classification codes can be made a monitoring search formula .

It is possible to extract at least one or more domestic and foreign inventors who apply for or register the largest number of patents in at least one or more of the most preferable patent classification codes,

(Backward citation, forward citation) are extracted from the document set to be managed by the unit patent information service system of the applicant's name or the unit patent information service system of the inventor's name, And / or the most probable applicant can be extracted, and the most probable applicant or applicant can be used as the applicant for the monitoring search formula.

In order to monitor a person who cites one or more individual patent documents of the document set (in order to find a person citing the patent document), the application number or registration number of the entire patent document constituting the document set is grouped into an OR operator It can be done by a search expression.

It is preferable that the monitoring search formula of the whole document set unit managed by the unit patent information service system of the applicant's name or the unit patent information service system of the inventor is folded for each monitoring subject and exposed to the user while being present under the monitoring service. For example, it is preferable that a folder named "Patent surveillance through the best patent classification code" is created in the monitoring subject of the surveillance search formula of the first method. In the folder, It would be desirable for the structure to be made. It is preferable that the result monitored by the monitoring search formula of the entire document set unit is confirmed in the folder or the hierarchical structure. Of course, the search formula to be set in advance can be modified by the manager or the inventor of the unit patent information service system of the applicant or the unit patent information service system of the inventor.

Patent intelligence module (60)

Hereinafter, the patent intelligence module 60 of the present invention will be described. The patent intelligence module 60 of the present invention performs a high-level advanced patent analysis, stores the result of the advanced patent analysis, or provides the user with the result of the advanced patent analysis.

The patent intelligence module 60 includes an analysis module 601 including a citation analysis module 601-1 and a comparison analysis module 601-2, a core field discovery module 602 related to discovery of a core field, A relatedness finding module 603 for generating collective information, a document collective attribute discovery module 604 for finding collective attributes of document aggregation units such as the degree of red ocean, a discovery module 604 for discovering a new subject, A fusion analysis module 606, a prospective field analysis module 607 for finding a promising field, a technology analysis module 606 for finding a blank technology, and a patent strategy analyzing a patent strategy. And a module 609. [0060]

The core field discovery module 602 includes a core applicant discovery module 602-1 for discovering a core applicant, a core inventor discovery module 602-2, a core agent discovery module 602-3, (602-4).

The relatedness finding module 603 includes a related applicant discovery module 603-1, a related inventor discovery module 603-2, a related agent discovery module 603-3, and a related patent classification symbol discovery module.

The document set collective attribute discovery module 604 includes a red ocean attribute discovery module 604-1.

The new subject discovery module 605-1 includes a new entry applicant discovery module 605-1 and a new subject discovery module 605-1. -1, the new entry inventor detection module 605-1-2, the new entry agent detection module 605-1-3, and the selection subject discovery module 605-2 are provided with a selection applicant discovery module 605-2- 1, the inventor finding module 605-2-2, and the selecting agent discovery module 605-2-3.

The fusion analysis module 606 includes a subject fusion fusion pattern analysis module 606-1, a technology fusion fusion pattern analysis module 606-2, a country fusion fusion pattern analysis module 606-3, And a fusion pattern analysis module 606-4. The generation module of the multi-patent classification symbol preliminary processing module 3900 and the multidimensional analysis operation execution result table 205 -7 for fusion analysis of the present invention and the fusion module 606 in the analysis module using the module Respectively. In the case where the technical or technical field is limited to patent classification symbols such as IPC, USPC, FT and the like, from the viewpoint of the patent classification symbol and the patent classification symbol per applicant / inventor, And explained that it is possible to limit the timing for each of the above.

The prospective field analysis module 607 includes a prospective field analysis module 607-1 and a prospective fusion field analysis module 607-2. The patent strategy analysis module 609 includes a patent strategy evaluation index analysis module 609-1, a patent utilization analysis module 609-2, and an acquired patent analysis module 609-3.

Unit patent information service system of applicant's name

As an application example of the patent information system of the present invention, a hierarchically integrated patent information service system and an individual unit patent information service system will be described.

The unit patent information service system of the applicant of the present invention is a patent information system in units of individual applicants, and the patent document information included in the unit patent information service system of the applicant includes the name of the applicant in the applicant and / Processing data such as various analyzes, statistical processing, information extraction, and the like of the present invention described above are generated in advance or in real time on the specified set of patent documents, And is provided to a computer. The inventor's unit patent information service system refers to a patent information system in units of individual inventor's name. The inventor's unit patent information service system includes a patent document set related to a specific inventor, Statistical processing, contract extraction, and the like are generated in advance or in real time and provided to the user's computer.

 It goes without saying that at least one unit patent information service system of the inventor's name can correspond to the unit patent information service system of one applicant. The unit patent information service system of the inventor may be configured under the unit patent information service system of the applicant. At this time, the manager of the unit patent information service system of the applicant, You will be able to manage access to the system.

The detailed configuration of the unit patent information service system of the individual unit applicant of the present invention is shown in Fig. In the unit patent information service system of the individual unit, the unit patent information service system generation engine of the applicant's name, which generates the unit patent information service system of the applicant, the unit patent information service system name of the inventor who creates the patent information system of each inventor An engine, a management module for managing the unit patent information service system of the applicant's name and / or the unit patent information service system of the inventor, unit patent information data of applicants managed in units of applicants, unit patent information of the inventor name There is data.

In the unit patent information service system generation engine 1100 of the patent owner, when the applicant is determined, the unit document acquisition module 1110 which obtains the document to at least one country having the applicant / patentee as the applicant, A unit document set DB generation module 1130, which is an application name that generates information on the document of the unit of the applicant's name in the data (including the table or DB or view), and the statistical analysis, A unit patent information intelligence generation module 1150 of the applicant who generates the intelligence information of the present invention such as monitoring, reporting and the like.

The inventor's unit patent information service system generation engine 1300 of the present invention operates in the same manner as the unit patent information service system generation engine of the applicant and includes an inventor's unit document acquisition module 1310, (1330) and an inventor's unit patent information intelligence generation module (1350).

In the management module 1500 of the present invention, the unit management module 1510 of the name of the applicant who manages the unit patent information service system of the applicant, the unit management module of the name of the inventor managing the unit patent information system of the inventor And a management information DB 1550 including the management information DB 1550. The unit management module of the applicant's name includes a unit document set management module 1511 of the applicant who manages the document set list of the applicant's name unit, a unit patent information intelligence management module 1513 of the applicant who manages the patent information intelligence unit of the applicant's name, There is an inventor's rights management module 1515 for managing the rights of the individual inventors belonging to the unit of the applicant's name, and a multi-tier hierarchical generation management module 1517 for managing the inventors and the like as a multi-tier hierarchy with at least one step. Meanwhile, the inventor's unit management module 1530 includes a unit document set management module 1531, an inventor's unit patent information intelligence management module 1533,

On the other hand, the unit patent information data 1700 of the applicant includes unit patent document set DB 1710 of applicant's name and unit patent information intelligence DB 1730 of applicant's name. In the inventor's unit patent document mast DB 1900, A patent document set DB 1910 and an inventor unit patent information intelligence DB 1930.

Meanwhile, a unit patent information service system of an individual applicant may have a unit patent information service system of at least one other applicant, and a method of having the unit patent information service system of another applicant may be a unit patent information service system of another applicant , Or there may be a method in which the unit patent information service system of another applicant is under his / her management authority. Having the unit patent information service system of another applicant as described above is also possible for an individual inventor who is the subject of the inventor's unit patent information service system.

The individual unit patent information service system generation engine of the present invention includes at least one of the unit patent information service system generation engine of the applicant's name and the unit patent information service system generation engine of the inventor's name. The unit patent information service system generation engine of the applicant includes a unit document set acquisition module of the applicant's name and a unit document set management module of the applicant's name. The unit patent information service system generation engine of the inventor's name includes an extraction module of the name of the inventor, An aggregate acquisition module and an inventor's unit document aggregation management module. The unit patent information service system generation engine of the applicant of the present invention and the unit patent information service system generation engine of the inventor have been registered in the patent database for each country by at least one country when the name of the applicant is obtained, It is possible to obtain all patent documents, automatically extract the names of individual inventors from the obtained patent documents, automatically acquire the patent documents for each individual inventor name, and obtain a set of patent documents in units of the applicant's name and individual inventor's units under the applicant's name A set of patent documents of the applicant, and management of the set of patent documents of the inventor

The unit patent information service system of the applicant of the present invention includes a set of patent documents related to the applicant's name. The patent document set includes a document set of the applicant's name registered or registered in the applicant's name and a document set of the inventor's name per individual inventor included in the document set of the applicant's name. The set of unit documents of the applicant and the set of unit documents of the inventor name may be obtained from the entire patent DB including at least one country patent DB, but preferably from the patent database of each country. In order to achieve this, if the applicant's identification code is unified for each country (the applicant's representative name is not only based on the first country but also on the basis of at least one second country, If the applicant of the second country is the same subject, the same applicant's identification code may be assigned. The applicant is preferably represented by at least one of the first country standard or the second country standard.

Through the above process, a partial patent document subset belonging to a unit of a specific applicant can be extracted, and a unit patent information service system in the name of the applicant having a subset of the patent document as a management target document can be created.

76 shows an exemplary process of the CRUD executed by the unit patent information service system administrator of the applicant on a document set managed by the unit patent information service system administrator.

The management module permits the administrator having the management authority on the unit patent information service system of the applicant's name (S2620). When there is a change in the management target patent document set by the administrator, the change module obtains the change in the set (S2630 If there is a change in the unit patent information service system of the inventor name by the manager (S2640), the change information is obtained in step S2640. If there is a change in the unit patent information service system of the second applicant by the manager (S2650). When there is a change in the service provided by the unit patent information service system of the applicant by the manager (S2660), the change information is obtained (S2660) and the obtained change information is stored (S2670) .

FIG. 77 shows an exemplary process of the CRUD performed by the unit patent information service system administrator of the inventor on a document set managed by the unit patent information service system administrator.

The management module permits the administrator having the management authority of the unit patent information service system in the name of the inventor (S2720). When there is a change in the management target patent document set by the manager, the management module obtains the change in the management target patent document set (S2730 If there is a change in the unit patent information service system of the inventor's name by the manager (S2740), if there is a change in the unit patent information service system of the second applicant by the manager, (S2750). If there is a change in the service provided by the unit patent information service system of the applicant by the manager (S2760), the change is obtained (S2760) and the obtained change is stored (S2770) .

The method by which the unit patent information service system generation engine of the inventor operates is the same as that of the unit patent information service system generation engine of the applicant, except that the target patent document set is specified by the inventor.

Extracting documents by inventor

If there is a whole list of inventors created by the extracting module of the inventor's name, the inventor and the inventor extract a subset of patent documents listed as inventors.

FIG. 66 is a flowchart of a process of creating a unit patent information service system of a single applicant in the patent information data of all applicants and generating a unit patent information service system of the inventor name in each inventor included in a set of patent documents constituting the unit patent information service system of the applicant And a method of performing the above method.

In step S1620, the unit patent information service system generation engine of the registered name searches the patent document registered or registered in the name of the acquired applicant through a search engine or a DB query (S1630) (S1640), it is determined whether there is a document to be registered or registered (S1640). Then, the unit patent information service system generation engine of the inventor extracts the inventors from the searched set of patent documents, and corresponds to the patent documents for each inventor (S1660). Then, the patent documents corresponding to the extracted inventors are set as a management target document set A unit patent information service system of the inventor name is created (S1670). If there is no document, the patent information service system of the applicant's name is not created.

67 shows a list of applicant names, generates a unit patent information service system of applicants by the number of applicants in the list from the unit patent information data of all applicants for each list, And shows an embodiment of a method of generating a unit patent information service system in the name of the inventor by the inventor included in the set of patent documents constituting the system.

The unit patent information service system generation engine of the patent application obtains the registered information of the applicant's name (S1720), acquires the information of the name of the individual application in the list (S1730), obtains the registered patent document (S1740). If it is judged that there is a document (S1750), a unit patent information service system of an individual applicant's name, which uses the searched patent document as a management target document set, is created (S1760) The information service system generation engine extracts the inventor from the set of patent documents, and corresponds to the patent document for each inventor (S1770). Then, the unit patent information service system of the inventor name, which sets the patent document corresponding to each extracted inventor as the management target document set, is created (S1780), the above process is not performed on the list of the acquired application name (S1790), the unit patent information service system of the applicant's name and the unit patent information service system of the inventor's name are repeated until there is no next applicant.

68 is a view for obtaining a list of applicants from the acquired document set, generating a unit patent information service system of the applicant's name as the number of applicants in the list from the unit patent information data of the entire applicant for each list, An inventor's unit patent information service system included in a set of patent documents constituting a unit patent information service system is generated according to an embodiment of the present invention.

The unit patent information service system generation engine obtains a document set for creating a unit patent information service system in the applicant's name (S1820), extracts the applicant's name from the document set, and generates the applicant's name list information (S1830)

In step S1840, the unit-patent information service system generation engine obtains the registered information of the applicant's name and obtains the information of the individual applicant's name in the list (S1840). The patented information service system generation engine searches the registered or registered patent document (S1850). If it is determined that there is a document (S1860), the unit patent information service system of the individual applicant's name is searched (S1870) and the unit patent information service The system creation engine extracts the inventor from the set of patent documents, responds to the patent document for each inventor (S1880), and creates the unit patent information service system of the inventor's name in which the patent document corresponding to each extracted inventor is a management target document set S1890), the above-described process is applied to the list of the acquired application name Determining whether the holder did not filed and until there is (S1891) and repeats the above patent information services unit of the applicant who created the system and inventors who patent information service system units produced.

69 is a flowchart showing a process of generating a unit patent information service system of a single applicant from the unit patent information data of all the applicants when the generation of the unit patent information service system of the applicant is ordered, And shows an embodiment of a method of generating a unit patent information service system in the name of the inventor by the inventor included in the set.

In step S1920, the unit patent information service system generation engine of the patent application obtains the application owner information from the external third party (S1920), searches the patent document registered or registered in the name of the acquired applicant through the search engine or the DB query (S1930). If it is determined that there is a document registered or registered in the name of the applicant (S1940), the unit patent information service system in the name of the applicant is searched (S1950). Then, the unit patent information service system generation engine of the inventor extracts the inventors from the searched set of patent documents and corresponds to the patent documents for each inventor (S1960). Then, the patent documents corresponding to the extracted inventors are set as a management target document set A unit patent information service system of the inventor name is created (S1970). If there is no document, the patent information service system of the applicant's name is not created.

The method of FIGS. 66 to 69 may be applied to both cases where the individual country is not specified or the individual country is specified. If it does not specify an individual country, it will be clear that the applicant's representative name should be presupposed (especially if the language is different).

FIG. 70 relates to a method for generating a unit patent information service system in the name of applicant in a country unit. In order to create a unit patent information service system of the applicant's name in this country unit, it is necessary to create unit patent information service system of unit applicant's name and unit patent information service system of the inventor's name, . Therefore, the methods of FIGS. 66 to 69 can be applied to all of them.

The unit patent information service system generation engine of the patent application obtains the information of the applicant's name (S2020), obtains the country list (S2030), and registers the patent document registered or registered in the name of the applicant (S2050). If it is determined that there is a document (S2050), the unit patent information service is completed in the country unit. If there is a patent document found in the country unit, (S2060). Then, the unit patent information service system generation engine of the inventor extracts the inventors from the set of patent documents at national level, and copes with patent documents for each inventor (S2070). Then, The inventor's name unit patent information service (S2080), it is judged whether there is a country that has not been processed in the country list (S2090). In the case where there is no such country, the patented document registered or registered in the name of the applicant The step of searching (S2040) is performed again.

FIG. 71 shows an example of a process of generating a unit patent information service system of an applicant's name and a unit patent information service system of an inventor in a country unit by utilizing family information.

In step S2120, the unit-patent information service system generation engine of the patent application obtains the information of the applicant's name (S2120), searches for the patent document registered or registered in the name of the applicant acquired in the first country (S2130) S240). If not, the unit patent information service system of the first country's standard applicant name is created (S2150), and the unit patent information system The patent information service system generation engine extracts the inventors from the set of patent documents retrieved from the first station, and maps the patent documents to each inventor (S2160). The patent document corresponding to each inventor extracted from the first country is set as a management target document set (S2170). Then, the unit patent information system The information service system generation engine obtains the family patent number in the second country based on the patent application number of the patent document searched in the first country (S2180), and obtains the patent number that can be obtained as the family patent number in the second country (S2190). Then, the unit patent information service system generation engine of the inventor's name sets the patent document set obtained from the second country to the inventor's patent information service system (S2191), and creates a unit patent information service system in the name of a second-party standard inventor in which the patent document corresponding to each inventor extracted from the second country is a management target document set S2193),

72 shows an embodiment of a process of generating a unit patent information service system of an applicant's name and a unit patent information service system of an inventor in a country unit by using a priority claim number.

In step S2220, the unit patent information service system generation engine of the patent application obtains the information of the applicant's name (S2220), searching for a patent document registered or registered in the name of the applicant acquired in the first country (S2230) (Step S2240). If there is no document, the first patent information service system of the first country, which is completed in the first country and, if there is a patent document searched in the first country, is set as the management target document set (S2250) The unit patent information service system generation engine of the inventor's name extracts the inventors from the set of patent documents retrieved from the first station, and corresponds to the patent documents for each inventor (S2260). The patent documents corresponding to the inventors extracted from the first station are managed A unit patent information service system of the first inventor name of the first country is created (S2270).

Then, the unit patent information service system generation engine of the patent application searches the list of patent application numbers of the patent documents retrieved from the first station from the priority claim number in the second station (S2280) (S2290). Then, the unit patent information service system generation engine of the inventor's name is obtained from the second country, which is obtained from the second country (S2291). In the case of a unit patent information service system in the name of a second-party standard inventor, which sets the patent document corresponding to each inventor extracted from the second country as a management target document set, extracts the inventor from the set of patent documents, (S2293).

FIG. 73 shows an example of a process of generating a unit patent information service system of an applicant's name and a unit patent information service system of an inventor in a country unit by utilizing family information.

In step S2320, the unit patent information service system generation engine of the patent application obtains the registered patent information (S2320), and searches for a patent document registered or registered in the name of the applicant acquired in the first country (S2330) (S2340). If there is the unit patent information service system, the unit patent information service system of the first country reference applicant is created (S2350). The unit patent information service system generation engine of the inventor's name The inventor is extracted from the set of patent documents retrieved from the first country, the patent document is corresponded for each inventor (S2360), and the patent document corresponding to each inventor extracted from the first country is referred to as the first inventor The unit patent information service system is generated (S2370)

The unit patent information service system generation engine of the patent application obtains the family patent number in the second country based on the patent application number of the patent document searched in the first country (S2380), and obtains the family patent number (S2390) a patent information service system in the second country in which the patent document that can be obtained is set as a management target document (S2390), and the patent document obtained from the second country is referred to as the first patent information service system (S2391), and then the unit patent information service system generation engine of the inventor's name is integrated with the patent document obtained from the integrated second country (S2393), which is a unit patent information service system in the name of the second inventor.

74 shows an example of a process of generating a unit patent information service system of an applicant's name and a unit patent information service system of an inventor in a country unit by using family information and priority claim information.

The unit patent information service system generation engine obtains the information of the applicant's name (S2420), searches for a patent document registered or registered in the name of the applicant acquired in the first country (S2403), and determines whether there is a document Next, if there is a unit patent information service system, the unit patent information service system of the first country reference applicant is created (S2450), and the unit patent information service system generation engine of the inventor's name The inventor is extracted from the set of patent documents retrieved from the first country, the patent document is corresponded to each inventor (S2460), and the name of the first inventor of the first country, which makes the patent document corresponding to each inventor extracted from the first country, The unit patent information service system is created (S2470).

Then, the unit patent information service system generation engine of the patent application searches the list of patent application numbers of patent documents searched in the first station from the priority claim number in the second station (S2480) The unit patent information service system generation engine of the inventor's name is acquired from the second country (step S2490) The patent document is integrated into the unit patent information service system of the first inventor name based on the patent application number retrieved from the first country via the priority claim number (S2491), the patent document obtained from the integrated second country (Step S2493), which is a unit patent information service system of the second inventor name of the second country, which is a management target document set.

FIG. 75 shows an exemplary process of the applicant's unit patent information service system generation engine and inventor's unit patent information service system generation engine when a new document is added.

The unit patent information service system generation engine of the patent application obtains a new patent document (S2620), extracts the applicant information from the new patent document (S2630), and judges whether there is a unit patent information service system in the name of the existing applicant with the applicant information (S2640).

If the judgment result is YES, the new patent information information is added to the unit patent information service system of the existing applicant (S2650). Then, the unit patent information service system generation engine of the inventor's name extracts the inventor information (S2660) If it is determined that there is an existing inventor's unit patent information service system in the lower part of the patent information service system (S2680), new patent document information is added to the unit patent information service system of the existing inventor (S2690) The unit patent information service system of the inventor's name is created based on the inventor's information, and the unit patent information service system of the inventor's name is created (S2699) with the new patent document information as the management target document set.

On the other hand, if there is no patent information service system of an existing applicant with applicant information, the unit patent information service system generation engine of the applicant's name generates a unit patent information service system of applicant's name with a new patent document as a management target document set (S2693 Then, the unit patent information service system generation engine of the inventor extracts the inventor from the new patent document and makes the patent document correspond to each inventor (S2695). Then, the inventor who sets the patent document corresponding to each extracted inventor as the management target document set Unit patent information service system is created (S2697).

Unit analysis of applicant's name or inventor's name

The at least one analysis algorithm set in the analysis module of the present invention is pre-set in the unit patent information service system of the applicant and / or the unit patent information service system of the inventor. That is, when the unit patent information service system of the applicant's name or the unit patent information service system of the inventor is created or after the creation, the analysis algorithm required in units of the applicant's name or the analysis algorithm Is automatically set to constitute the analysis service of the unit patent information service system of the applicant's name or the unit patent information service system of the inventor's name. The execution of the analysis algorithm is executed when generating the unit patent information service system of the applicant's name or the unit patent information service system of the inventor's name and outputs the analysis result to the unit patent information service system of the applicant's name or the unit patent information service system It is preferable to have this in advance. However, if the execution speed of the analysis algorithm is fast, or is fast enough to not give real time or waiting for the user, the analysis algorithm may be executed when there is external access (such as clicking of an administrator or a user) There will be. Therefore, when the analysis module is configured with at least two analysis algorithms, the analysis subject and the analysis algorithm thereof are set in advance, and the execution time thereof may be different.

Management module

Hereinafter, the multi-layer management module included in the management module will be described. First, a description will be given of a situation in which a problem of multi-layer management occurs. If an applicant has a considerably large number of inventors, the inventor may belong to at least one hierarchical organization. For example, one inventor may belong to the a1 team <a department <A1 series <A company <alpha group. Therefore, in the management aspect, it is natural that the a1 team manages the patent documents related to the patent information service system and the a1 team of the inventor unit, and the a department is a unit of the inventor's name related to each department from a1 to an The patent information service system and / or the patent document are subject to management, the A1 series covers the patent information service system and / or the patent document of the inventor's name related to the department a to the z department, The alpha group is a unit patent information service system in the name of the applicant from A company to Z company, a unit patent information service system in the name of the inventor, and / or a patent document, etc. Will manage it. However, since most of the information available in the published document is only the inventor's name and company name (that is, the patent document usually contains no information on the group to which the inventor belongs in addition to the applicant's information) The manager and / or the highest level manager need to be able to organize where the specific patent document should belong. Since the patent information service system is generated through a patent document to be managed, when there is information on belonging to which inventor, which team, which department, which series, which company the patent document corresponds to, The patent information service system can be created. In order to determine the affiliation information, information about the patent document, the inventor, and the company can be obtained from the patent document itself. Therefore, it is possible to know whether the inventor belongs to which team, which department, which series, which company, or which team, which department, which series, which company the document itself belongs to, , The group to which the patent document belongs can be determined.

Individual unit patent information system management module

When two or more applicants' unit patent information systems are created, the patent information service systems need to be managed, and the management is performed by the open patent information service system management module. The open patent information service system management module manages authentication, authority management, transfer of authority, deletion, change of location and affiliation of the unit patent information service system of the applicant.

When the unit patent information service systems of a plurality of applicants are created as described above, it is necessary to integrally manage the patent information service systems from the viewpoint of the entire system, and the integrated management module of the present invention performs the integrated management. FIG. 65 shows an embodiment of the configuration of the integrated management module. The integrated management module includes an individual unit patent information system multi-stage grouping module 6100 and an individual unit patent information system management module 6300, The information service system can be multi-tiered, and the multi-tier grouped patent information service system can be managed.

Implementation of Applicant's Patent Information Service System

124 is an exemplary embodiment of the 3COM patent information system, which is selected when 3COM (No. 6) is selected as one of Nasdaq-registered companies, when the US patent tab is selected in the patent list of the patent portfolio among the patent portfolios . When there is an input to the 3COM, the patent information service system of the present invention can retrieve US documents for 3COM or query the patent document master DB to generate a document list for 3COM and present it to the user.

125 is a screen in which a list of inventors of 3COM is displayed when the inventor list is selected in FIG. 124 and the United States is selected in the countries tab. The patent information service system of the present invention can retrieve all the inventor's lists of 3COM by querying the patent document master DB or querying the operation result table of the inventor's analysis multi-dimensional analysis operation. On the other hand, the inventors of 3COM may be listed in ABC order, but it will be obvious to those skilled in the art to arrange them in different order such as multi-application / multi-registration or multi-user.

FIG. 126 is a screen listing a patent document related to the inventor when Aldous Stepha .. (No. 9) is selected from the inventor list in FIG. 125. FIG. The patent information service system of the present invention can obtain selection information for the inventor and then search the patent document master DB with the acquired inventor information to create a list of documents corresponding to the inventor and present it to the user will be.

Figure 127 shows the number of applications for each IPC per year at the IPC subclass level based on the total amount of applications when the United States is clicked in the systematic analysis menu (SA), the United States by pressing the statistical analysis tab in Figure 124 for 3COM, Is an example embodiment. This screen pops up a simple analysis screen that appears when you drill down and press a specific cell value.

FIG. 128 shows the results of the competition based on the number of applications per USPC per year at the USPC subclass level based on the total amount of applications when the user clicks on the statistical analysis tab in FIG. 124 for 3COM and the US systematic analysis menu This is an exemplary embodiment in which information on the applicant comes out.

129 shows a case where the statistical analysis tab of FIG. 124 for 3COM is clicked and the entire set of application documents of the 3COM filed when the US is clicked on the systematic analysis menu (SA) and the country tab is set as a citation reference document set, (backward) citation document set as an analysis target document set.

130 is a graphical representation of the results of the inventor's systematic analysis menu (SA) for the 3COM, the statistical analysis tab of FIG. 124, the number of applicants for each year by IPC at the IPC subclass level based on the total amount of applications, Is an example embodiment.

FIG. 131 is a diagram of an exemplary method of automatically generating processed patent information for a set of patent documents by automatically including information on a sub-patent classification.

132 is a diagram of an exemplary method for generating processed patent information, including multidimensional analysis, when a set of patent documents is selected.

133 is a diagram of an exemplary method for performing a fusion analysis when a set of patent documents is selected. One example of the result of performing the fusion analysis is shown in FIG. 142 to FIG. The result of performing the fusion analysis may be represented by a network diagram. That is, the applicant, the inventor, the patent classification code, the representative keyword, the technical keyword, the patent document number (referring to n patent references included in one patent document) constitute nodes on the network diagram, And a numerical value in the case where a co-occurrence occurs between the node and the node. The size of the node may reflect the number of the document or other indicator. Needless to say, if a user clicks on a number, the list of the document set can be displayed as shown in FIG. In order to generate the network diagram, the association result data must be generated between the nodes extracted from the specified patent document set.

134 is a diagram of an exemplary method for providing patent document information corresponding to a numerical value, when there is a selection for the numerical value; A concrete example of this can be seen in FIG. In Figure 48, a pop-up window opens when a number is pressed, showing that the patent list is provided.

135 is a diagram of an exemplary method for providing multi-dimensional analysis by applicant, patent classification, inventor, and agent. 17 to 50 of the present invention illustratively show that there are various menus for each analysis type.

On the other hand, even if a patent is filed or registered in the first country under the principle of patent independence, it is not protected in the second country unless it is filed or registered in the second country. Therefore, it is possible to check whether there is family information related to the individual patent documents in at least one or more second countries with respect to the individual patent documents constituting the patent list from the first country patent search result. If there is no investigation result, it will be possible to check whether the individual patent document is internationally filed, and if it is an international application, And whether or not one year has elapsed since the filing of the first international application. When the inquiry result has elapsed, it is not possible to enter the second station, so that the information about the availability in the second station can be included in the predetermined vicinity of the search result of the first station patent. Of course, even if you enter a second country, information on availability can be provided even if the current status information is withdrawn, invalidated, rejected, or disappears. On the other hand, if the applicant enters the second country, the PCT is pending, does not exceed the predetermined period, or has not passed one year since the filing date of the first country, the risk / warning / It will be possible. This will be particularly useful when the system 10 knows that the user is a user belonging to the second station in the user's personal information and furthermore when the user designates the second stations he is interested in, It will be useful. The second country may be, of course, more than one country. 136 is a diagram of an exemplary method for determining the availability of patents retrieved from a first country in a second country using family information. FIG. 137 is a flowchart illustrating an example method for determining whether the method of FIG. 136 is JP as a first station and that the use of KR, US, and EU as a first station for a # 1 patent document in a second station FIG. 138 is an exemplary diagram of a search result list screen when the method shown in Figs. 136 to 137 is carried out. It can be seen that each item in the search result list includes information on the utilization level at the national level for each country.

One example of the monitoring service or the user-created directory service is the services shown in FIG. 139 to FIG. The user can construct a tree structure of the technical system / product structure / competitor that he / she is interested in, and can store related information such as a search expression or other recipient e-mail in each node constituting the tree. At this time, the system 10 may inquire the search formula when a predetermined period or a new patent document is obtained, and if the patent document is found, the system 10 may send the patent document by e-mail or the like. At this time, the recipient can send out the desired language using the machine translation system. Meanwhile, if a retrieval expression or a document set is allocated to the user-created directory, a set of patent documents can be specified for each node constituting the directory, and various analyzes of the present invention for a document set can also be performed for each node . Of course, two or more nodes may be selected, and various analyzes may be performed on a new set of patent documents generated by performing an aggregation operation between the nodes.

FIG. 139 is an exemplary embodiment of a surveillance service / directory service method of the present invention in which a description tree is configured in a multi-level hierarchical structure for each user and various information including a search formula is included in a description tree.

144 to 145 are examples of interfaces used to implement the patent information system of the present invention. The interface has many interfaces associated with the selection of options. The interface includes: 1) a time type selection interface, 2) a time interval selection interface, 3) a country selection interface, 4) a patent classification symbol selection interface, 5) a citation specialized interface, 6) And any one or more of them.

Wherein the time type interface comprises an interface providing a selection of at least one of a filing date, a publication date, and a registration date, wherein the time period selection interface is configured to select one or more of year / month / week / Wherein the country selection interface includes an interface for providing at least one selection for at least two countries, wherein the patent classification symbol selection interface includes a classification selection type interface for a patent classification symbol, An interface providing a selection of one or more of a selection interface, a main patent classification symbol, and a subpublic classifier selection interface, wherein the citation specialized interface includes a citation direction interface for selection of at least one of a front citation and a rear citation, An interface for providing a selection of at least one of a citation depth interface for selection of depth, an interface for selection of one or more citations between direct citations and indirect citations, wherein the family information containing selection interface includes: And an interface for providing a selection of one or more of the interfaces relating to the selection of the range.

Wherein the post-search selection interface presents a pre-set field search with the target document set as a search target, receives a search word from the user's computer in the field, provides a search result corresponding to the search word, And an interface that receives at least one selection information and processes only a set of patent documents related to the selection information as a set of object documents to be processed. For example, if only a specific applicant is targeted in a specified set of patent documents, if the applicant is entered in the applicant search field, a list of applicants included in the specified set of patent documents comes out, , The selection information is received, and only the selected document set is selected, or the selected document set is excluded, so that the respecification of the set of patent documents is completed.

The present invention can be applied to the entire patent information industry. Specifically, it can be applied to various patent information related industries such as search, analysis, consulting, surveillance, analysis of competitors, and technology fusion pattern research.

BRIEF DESCRIPTION OF DRAWINGS FIG. 1 is a diagram showing an embodiment of a patent information system according to the present invention. FIG.

2 is a diagram showing an embodiment of the DB unit of the present invention.

Fig. 3 is an embodiment of the patent document mast DB of the present invention.

Fig. 4 is a diagram showing an embodiment of the patent classification code mast DB of the present invention.

5 is a diagram showing an embodiment of the subject mast DB of the present invention.

FIG. 6 is a diagram illustrating an operation result table DB for multidimensional analysis of the present invention.

FIG. 7 is an exemplary diagram of a support DB part and a secondary processing DB part of the present invention. FIG.

8 is a diagram illustrating an embodiment of the pretreatment module unit of the present invention.

9 is a diagram showing an embodiment of a mast DB generating module according to the present invention.

10 is a diagram illustrating an embodiment of a citation information preprocessing module of the present invention.

11 is a diagram showing an embodiment of a patent information processing basic module of the present invention.

12 is an exemplary diagram of a support module of the present invention.

13 is an exemplary diagram of a patent intelligence module of the present invention.

Fig. 14 is an exemplary diagram of the delphion of the present invention relating to the reason why the sub-classification-containing search of the present invention is not analyzed according to the present invention.

15 is an exemplary diagram of an embodiment of the directory creation module of the present invention.

16 is an exemplary diagram of an embodiment in which the patent classification code search module of the present invention is operated.

FIG. 17 is a diagram showing an embodiment analysis result of the number of applications per year for each multi-application IPC at the IPC subclass level of Samsung Electronics Co., Ltd. filed in the Republic of Korea of the present invention.

18 is an exemplary diagram of the analysis result generated when the user drills down to H01L.

19 is an exemplary diagram of the analysis results generated when drilling down to H01L 21/00 and other subpublic classifications.

FIG. 20 is a diagram showing an exemplary analysis result of application number data per year of IPC main group level of Samsung Electronics Co., Ltd. among all applicants in the DB possessed by the patent information system of the present invention.

FIG. 21 is a diagram showing an exemplary analysis result of the number of applications per year by IPC at the IPC 1-dot subgroup level of Samsung Electronics Co., Ltd. among all applicants in the DB possessed by the patent information system of the present invention.

FIG. 22 is a graph showing the result of analysis of one embodiment of the number of applications per year by IPC subclass level based on application documents of Samsung Electronics Co., Ltd. filed in the United States among all applicants in the DB owned by the patent information system of the present invention FIG.

FIG. 23 is a diagram showing an embodiment analysis result of the number of applications per year for each multi-application IPC at the IPC main group level based on the registration document of Samsung Electronics Co., Ltd., registered in the US among all applicants in the DB owned by the patent information system of the present invention to be.

FIG. 24 is a graph showing the result of analysis of one embodiment of the number of applications per USPC according to the USPC no dot (subclass) level of General Motors filed in the US among all the applicants in the DB owned by the patent information system of the present invention FIG.

FIG. 25 is a diagram showing results of analysis of one embodiment of the application number data per year of multi-application IPC at USPC 1 dot level of General Motors, which was filed in the United States among all applicants in the DB possessed by the patent information system of the present invention.

26 is an exemplary diagram of an example of a total amount analysis and drill down for IPC H04B based on Korean application document of the present invention.

Figure 27 is an exemplary illustration of an example of a total volume analysis and drill down for IPC H04B based on the US application of the present invention.

FIG. 28 is a diagram showing an embodiment of the total amount analysis for each applicant company of IPC H04B based on the Korean application document of the present invention.

FIG. 29 is an exemplary view of applicants based on the occupancy rate of IPC H04B based on the Korean application document of the present invention.

Figure 30 is an embodiment of an applicant of the present invention based on the Korean application document based on IPC H04B.

Figure 31 is an exemplary diagram of a total volume analysis including a drill down to IPC H04B and their subclasses based on the US application of the present invention.

FIG. 32 is a diagram illustrating a representative competitor analysis based on the total amount of the Korean standard Samsung Electronics Co., Ltd. among all the applicants in the DB possessed by the patent information system of the present invention.

FIG. 33 is an exemplary diagram of an analysis on a competitive applicant according to the patent application technology classification number based on the total amount of all the applicants in the database owned by the patent information system of the present invention, Samsung Electronics Co., Ltd., Korea.

FIG. 34 is an exemplary diagram of an analysis of all the applicants in the DB held by the patent information system of the present invention to the competition applicant based on the IPC main group standard application total amount of US Patent Standard Samsung Electronics Co., Ltd. FIG.

FIG. 35 is an exemplary diagram of an analysis result of multi-filing inventors of SAMSUNG ELECTRONICS CO., LTD. Based on the total amount of Korean patent applications among all the applicants in the DB held by the patent information system of the present invention.

36 is an example of an analysis result of multi-application multi-applicant inventors of Samsung Electronics Co., Ltd. based on the total amount of Korean patent applications among all applicants in the DB possessed by the patent information system of the present invention.

FIG. 37 is a USPC subclass (based on the total amount of all US patent applications of Samsung Electronics Co., Ltd.) among all applicants in the DB held by the patent information system of the present invention. Ranking analysis Fig.

FIG. 38 is a graph showing the relationship between the backward citation document set of the present invention and the document set of the analysis when all the US patent applications of Samsung Electronics Co., Ltd. among all the applicants in the DB held by the patent information system of the present invention are set as reference documents. An example of the result of analysis of the total amount of citations by year.

FIG. 39 is a graph showing the relationship between the backward citation document set of the present invention and the document set for analysis when all the US patent applications of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system of the present invention are set as the reference document set. FIG. 4 is an exemplary diagram of an analysis result of an applicant of the present invention by year; FIG.

FIG. 40 is a graph showing the relationship between the backward quoted document set of the present invention and the document set for analysis when all the US patent applications of Samsung Electronics Co., Ltd., among all the applicants in the database held by the patent information system of the present invention, An IPC main group level patent classification code (IPC main group level).

FIG. 41 is a graph showing the relationship between the backward quoted document set of the present invention and the document set for analysis when all the US patent applications of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system of the present invention are set as the reference document set. And an analysis result for each year reflecting a case where a patent classification code (IPC main group level) is drilled down.

FIG. 42 is a graph showing the relationship between the backward quoted document set of the present invention and the document set of the analysis when all the US patent applications of Samsung Electronics Co., Ltd. among all the applicants in the DB held by the patent information system of the present invention are set as the reference document set. An embodiment of the result of yearly analysis of the inventor of the present invention.

FIG. 43 is a diagram showing a set of document specifications for analysis of the present invention when all US patent registration of US patent corporation among all applicants in the DB owned by the patent information system of the present invention is set as a reference document set. , And an example of the yearly analysis result for the most cited applicant.

FIG. 44 is a view showing a set of document specifications for analysis of the invention, which is a set of forward quotations of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. is set as a reference document set among all applicants in the DB possessed by the patent information system of the present invention, An embodiment of the result of yearly analysis of the most cited inventors.

FIG. 45 is a diagram showing a set of document specifications for analysis of the invention, which is a set of forward quotations of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. is set as a reference document set among all applicants in the DB possessed by the patent information system of the present invention, It is an exemplary diagram of the yearly analysis result when drilling down to the IPC among the analysis of the most cited technologies per IPC main group.

FIG. 46 is a diagram showing a set of reference documents of the present invention when all the US patent registration of Samsung Electronics Co., Ltd. is set as a reference document set among all applicants in the DB held by the patent information system of the present invention, It is an exemplary diagram of the yearly analysis result when the USPC subclass (no dot, subclass subclass) most quoted technology analysis is drill down.

47 is a diagram showing a set of reference documents related to the entire rear citation of the present invention when all US patent registration of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system of the present invention is set as a reference document set, And a chart generated by the chart generating module among the reporting modules of the present invention with respect to the analysis result.

FIG. 48 is a diagram showing a set of reference documents related to the entire rear citation according to the present invention when all US patent registration of Samsung Electronics Co., Ltd. among all applicants in the DB held by the patent information system of the present invention is set as a reference document set, The results of the analysis of the cited total amount of inventors who received a lot of citations and the results of simple analysis (document list, application by year of the largest applicant, Number of applications / number of registrations, number of applications / number of registrations by year, maximum number of applications by IPC, USPC, FT by year (including drill-down) Fig.

FIG. 49 is an exemplary diagram showing a drill down function according to the maximum technology field (IPC, USPC, FT) in the simplified analysis module of the present invention.

50 shows the backward citation document of the present invention when all of the US application documents of Samsung Electronics Co., Ltd. among all the applicants in the DB held by the patent information system of the present invention are documents of multi-application IPC subclass unit as a reference document set FIG. 5 is an exemplary diagram of the yearly analysis result on the total amount of citations, which is a set of documents to be analyzed.

51 is a diagram showing an embodiment of a hierarchical patent information service system which is one sub-system of the patent information system of the present invention.

52 is an exemplary diagram of an individual unit patent information system generation engine of the present invention.

53 is a diagram showing an internal configuration of a patent information preprocessing module of the present invention.

54 is a diagram illustrating an embodiment of the weight pre-processing module of the present invention.

55 is a diagram showing an embodiment of the citation information preprocessing module of the present invention.

56 is an exemplary diagram of the patent classification symbol preprocessing module 301-3-1 or 3500 of the present invention.

57 is a diagram showing an embodiment of a representative pre-processing module of the present invention.

FIG. 58 is a diagram showing an embodiment of the representative word extraction pre-processing module of the present invention. FIG.

59 is a diagram showing an embodiment of the family information preprocessing module of the present invention.

FIG. 60 is a diagram illustrating an embodiment of a plurality of patent classification symbol related preprocessing modules of the present invention. FIG.

61 is a diagram showing an embodiment of a statistical preprocessing module according to patent classification symbols of the present invention.

62 is an exemplary diagram of a patent information intelligence module of the present invention.

63 is a diagram showing an embodiment of the analysis module of the present invention.

FIG. 64 is an exemplary diagram of a patent information system batch generation engine of the present invention. FIG.

65 is a diagram showing an example of the configuration of the integrated management module of the present invention.

FIG. 66 is a diagram showing an example of a patent information service system according to the present invention. In FIG. 66, a unit patent information service system of a single applicant is created in the entire patent information database 2300, FIG. 4 shows one embodiment of the method for generating the data.

67 shows a list of the names of applicants, generates a unit patent information service system of the applicant by the number of applicants listed in the entire patent information database 2300 for each list, And shows an embodiment of a method of generating a unit patent information service system in the name of the inventor by the inventor included in the set of patent documents constituting the system.

68 shows a list of the applicants in the obtained document set, generates a unit patent information service system of the applicant's name as the number of applicants in the list in the entire patent information database 2300 for each list, An inventor's unit patent information service system included in a set of patent documents constituting a unit patent information service system is generated according to an embodiment of the present invention.

FIG. 69 shows an example of a patent information service system in which a unit patent information service system of a single applicant is created in the entire patent information database 2300 when the generation of the unit patent information service system by the applicant is ordered, And shows an embodiment of a method of generating a unit patent information service system in the name of the inventor by the inventor included in the set.

FIG. 70 relates to a method for generating a unit patent information service system in the name of applicant in a country unit. In order to create a unit patent information service system of the applicant in the unit of country, it is necessary to create a unit patent information service system of a unit name of a applicant and a unit patent information service system of a unit name of an inventor in a national patent information database 2300 . Therefore, the methods of FIGS. 66 to 69 can be applied to all of them.

FIG. 71 shows an example of a process of generating a unit patent information service system of an applicant's name and a unit patent information service system of an inventor in a country unit by utilizing family information.

72 shows an embodiment of a process of generating a unit patent information service system of an applicant's name and a unit patent information service system of an inventor in a country unit by using a priority claim number.

FIG. 73 shows an example of a process of generating a unit patent information service system of an applicant's name and a unit patent information service system of an inventor in a country unit by utilizing family information.

74 shows an example of a process of generating a unit patent information service system of an applicant's name and a unit patent information service system of an inventor in a country unit by using family information and priority claim information.

FIG. 75 shows an exemplary process of the applicant's unit patent information service system generation engine and inventor's unit patent information service system generation engine when a new document is added.

Fig. 76 is a diagram of an embodiment relating to the present invention. Fig.

FIG. 77 shows an exemplary process of the CRUD performed by the unit patent information service system administrator of the inventor on a document set managed by the unit patent information service system administrator.

78 is a diagram of an embodiment relating to the present invention.

79 is an exemplary diagram of a method in which the cost expenditure perspective weight pre-processing module 3311 of the present invention processes weights.

80 is an exemplary diagram of a way in which the citation-based weight pre-processing module 3313 of the present invention prepares a weight in terms of a citation.

81 is an exemplary diagram of a manner in which the conflict-of-interest weight pre-processing module 3315 of the present invention processes weights.

82 is an exemplary diagram of a manner in which the concentrated-point weighting preprocessing module 3317 of the present invention processes weights.

FIG. 83 is an exemplary diagram of a method for processing weights in the unit of an inventor of the present invention and processing the weights.

FIG. 84 is an exemplary diagram of a manner in which the unit weight pre-processing module 3331, which is an applicant of the present invention, processes weights.

85 is an exemplary diagram of a manner in which the inventor's unit weight pre-processing module 3333 of the present invention processes weights.

86 is an exemplary diagram of a manner in which the proxy unit weight preprocessing module 3335 of the present invention processes weights.

87 is an exemplary diagram of a method by which the family information preprocessing engine 3810 of the present invention processes family information.

88 is an exemplary diagram of a citation information preprocessing method for representing the citation number of the present invention.

FIG. 89 is an exemplary diagram of a method for incorporating the backward quotation document information into the document information of the specific document after acquiring the backward quotation document information, which is information on the later-filed document citing the specific document of the present invention .

90 is a diagram illustrating an embodiment of a citation information preprocessing method of the present invention.

91 is an exemplary diagram of a processing method of a patent classification symbol preprocessing engine for processing a modified patent classification code of the present invention.

FIG. 92 is an exemplary diagram of a hierarchical modified patent classification symbol database generation module of the present invention for generating the modified patent classification symbol database. FIG.

93 is a diagram of an exemplary method of making a tree structure with a USPC patent classification symbol.

94 is a diagram of an exemplary method of assigning a modified patent classification symbol corresponding to a tree structure to the USPC patent classification symbol of FIG. 93; FIG.

FIG. 95 is an exemplary diagram of a tree structure composed of modified patent classification symbols having the same structure as the tree structure formed by the USPC patent classification symbol of FIG. 93; FIG.

FIG. 96 is an exemplary diagram showing that a tree structure of a patent classification symbol as shown in FIG. 93 can be created with Index to US Patent Classification (aka, Classification Index File) file).

97 is a diagram showing an embodiment of a pretreatment method of a patent classification mark of the present invention.

FIG. 98 is an exemplary diagram of a method for performing the representative renormalization of the representative description pre-processing module of the present invention.

FIG. 99 is a diagram illustrating an exemplary embodiment of the present invention using a priority assertion number; FIG.

100 is a diagram illustrating an embodiment of a method for preprocessing statistical values for each patent classification symbol of the present invention.

101 is an exemplary diagram of a method for automatically generating a statistical value, parameter or calculated value by automatically including a subpublic classifier for a given patent classifier of the present invention.

102 shows a method for generating statistical values, parameter values, or calculated values according to patent classification symbols from the viewpoint of including a sub-patent classification symbol for citation or reference information on a document subset of a specific document set of the present invention Fig.

FIG. 103 is a diagram showing an embodiment of a method for preprocessing a plurality of patent classification symbols according to the present invention.

FIG. 104 is a diagram showing an embodiment of a plurality of patent classification symbol relationship preprocessing methods in a comparative view of the present invention. FIG.

105 is an exemplary diagram of a representative word information preprocessing method of the present invention.

106 is an exemplary diagram of a representative word information preprocessing method of the present invention.

107 is an exemplary diagram of a representative word information preprocessing method of the present invention.

FIG. 108 is a diagram illustrating an embodiment of an analysis index calculation method according to the present invention. FIG.

FIG. 109 is an exemplary diagram of a method for obtaining a set of patent documents analyzed according to the present invention. FIG.

110 is a diagram illustrating an embodiment of a trend analysis method of the present invention.

FIG. 111 is a diagram illustrating an embodiment of the simplified analysis method of the present invention. FIG.

112 is an exemplary diagram of a citation analysis method of the present invention.

113 is a diagram showing an embodiment of the analysis method for cited article of the present invention.

FIG. 114 is a diagram showing an embodiment of the method for analyzing the citation of the present invention. FIG.

115 is a diagram showing an embodiment of a method for analyzing a plurality of patent classification symbols of the present invention.

116 is an exemplary diagram of a method for analyzing a plurality of patent classification symbols of the present invention.

117 is a diagram illustrating an embodiment of a method for analyzing a plurality of patent classification symbols of the present invention.

118 is an exemplary diagram of a method of operating the individual unit patent information system multi-stage grouping module of the present invention.

119 is an exemplary diagram of a method for generating a patent information system batch of the present invention.

FIG. 120 is an exemplary embodiment of an applicant list screen for the top 500 applicant of the Republic of Korea as an embodiment of a screen on which the unit patent information system of the present invention is implemented. In the country tab, Korea is selected and the Top 500 tab is selected on the top 500 tab of the subtab, the registered company of the exchange, the registered company of the KOSDAQ, all of the applicants, and the whole company.

FIG. 121 is an exemplary embodiment of an applicant list screen for an Nasdaq-registered company in the United States, which is an embodiment of a screen on which the unit patent information system of the present invention is implemented. In the Countries tab, the United States is selected, and NASDAQ is selected in the Top 500, NYSE (New York Stock Exchange), NASDAQ, AMEX, There are a number of NASDAQ registered companies, so there are tabs in ABC order and the entire NASDAQ company tab.

FIG. 122 is an exemplary embodiment of an applicant list screen for a JASDAQ registered company in Japan as an embodiment of a screen on which the unit patent information system of the present invention is implemented. In the country tab, Japan is selected, and the top 500, JP1 (1 registered company of Tokyo Stock Exchange), JP2 (Tokyo Stock Exchange 2 registered company), JPM (MOTHERS registered company), JASDAQ registered company, JASDAQ in the screen is selected.

FIG. 123 is an exemplary embodiment of a list screen of applicants to a London Stock Exchange registered company in Europe as an embodiment of a screen on which the unit patent information system of the present invention is implemented. Europe is selected in the countries tab and UK1 is selected on the subtabs Top 500, UK1 (London stock exchange), AIM (London stock exchange), OVERSEAS LISTED (London foreign exchange listed companies), EURONEXT, The list of listed companies in Frankfurt is not shown.

FIG. 124 shows an example of an embodiment when a US patent tab is selected from the list of patents among the patent portfolios among 3COM's patent information systems that appear when 3COM (No. 6), which is one of NASDAQ registered companies in FIG. 121, This is an implementation example.

125 is a screen in which a list of inventors of 3COM is displayed when the inventor list is selected in FIG. 124 and the United States is selected in the countries tab.

FIG. 126 is a screen listing a patent document related to the inventor when Aldous Stepha .. (No. 9) is selected from the inventor list in FIG. 125. FIG.

Figure 127 shows the number of applications for each IPC per year at the IPC subclass level based on the total amount of applications when the United States is clicked in the systematic analysis menu (SA), the United States by pressing the statistical analysis tab in Figure 124 for 3COM, Is an example embodiment. This screen pops up a simple analysis screen that appears when you drill down and press a specific cell value. This is structurally equivalent to Figure 17, which is an implementation example for Samsung Electronics Co.,

FIG. 128 shows the results of the competition based on the number of applications per USPC per year at the USPC subclass level based on the total amount of applications when the user clicks on the statistical analysis tab in FIG. 124 for 3COM and the US systematic analysis menu This is an exemplary embodiment in which information on the applicant comes out. This screen is structurally equivalent to Fig. 38 with an implementation example for Samsung Electronics Co., Ltd. Fig.

129 shows a case where the statistical analysis tab of FIG. 124 for 3COM is clicked and the entire set of application documents of the 3COM filed when the US is clicked on the systematic analysis menu (SA) and the country tab is set as a citation reference document set, (backward) citation document set as an analysis target document set. This screen is structurally equivalent to Fig. 39, which is an implementation example of Samsung Electronics Co., Ltd. Fig.

130 is a graphical representation of the results of the inventor's systematic analysis menu (SA) for the 3COM, the statistical analysis tab of FIG. 124, the number of applicants for each year by IPC at the IPC subclass level based on the total amount of applications, Is an example embodiment. This is structurally equivalent to Fig. 35, which is an implementation example for Samsung Electronics Co., Ltd. Fig.

FIG. 131 is a diagram of an exemplary method of automatically generating processed patent information for a set of patent documents by automatically including information on a sub-patent classification.

132 is a diagram of an exemplary method for generating processed patent information, including multidimensional analysis, when a set of patent documents is selected.

133 is a diagram of an exemplary method for performing a fusion analysis when a set of patent documents is selected.

134 is a diagram of an exemplary method for providing patent document information corresponding to a numerical value, when there is a selection for the numerical value;

135 is a diagram of an exemplary method for providing multi-dimensional analysis by applicant, patent classification, inventor, and agent.

136 is a diagram of an exemplary method for determining the availability of patents retrieved from a first country in a second country using family information.

FIG. 137 is a flowchart illustrating an example method for determining whether the method of FIG. 136 is JP as a first station and that the use of KR, US, and EU as a first station for a # 1 patent document in a second station FIG.

138 is an exemplary diagram of a search result list screen when the method shown in Figs. 136 to 137 is carried out. It can be seen that each item in the search result list includes information on the utilization level at the national level for each country.

FIG. 139 is an exemplary embodiment of a surveillance service / directory service method of the present invention in which a description tree is configured in a multi-level hierarchical structure for each user and various information including a search formula is included in a description tree.

140 is an exemplary embodiment in which the method of FIG. 139 is applied to a product.

141 is an exemplary embodiment in which the method of FIG. 139 is applied to a company system of a competitor or the like.

142 is an exemplary embodiment in which a fusion analysis is performed on the inventor and the network diagram is implemented.

FIG. 143 is an exemplary embodiment in which a fusion analysis is performed on a technical keyword and a representative word, and is implemented as a network diagram.

144 to 145 are examples of interfaces used to implement the patent information system of the present invention.

Claims (8)

A system for providing patent convergence analysis information to a user's computer, Obtaining selection information on a set of patent documents; Specifying a set of patent documents to be processed based on the selection information; And generating processed patent information by performing at least one predetermined process on the specified set of patent documents to be processed, The method for receiving the selection information includes: 1) a method of receiving input from a user computer through a search formula; and 2) information constituting a bibliography included in the predetermined patent document presented on the user's computer from the user's computer A method for receiving selection information on a specific set of patent documents selected from among a set of at least one patent document previously generated by the user and 4) a method for receiving selection information for at least one A method of receiving a condition is used, and more than one method is used, Wherein the processed patent information is at least one or more multidimensional analysis information generated through at least one predetermined analysis process on the specified set of patent documents to be processed, The individual dimensions constituting the multidimensional are selected from among patent documents contained in 1) time, 2) patent classification symbol, 3) applicant, 4) inventor, 5) agent, 6) country, 7) Any one or more, The predetermined analysis process is an analysis process for a fusion pattern, Wherein the multidimensional analysis information includes at least one predetermined counting value for the at least one individual dimension, a calculation value by at least one predetermined calculation formula, a query by at least one quality consciousness, A result value, and a processing result value for at least one or more analysis processing algorithms, Wherein the system for providing patent information visualizes the result of the fusion analysis processing into a network diagram including at least two or more nodes, Wherein the visualization process further includes at least one of information about the node and information about a line connecting the node and the node. The method according to claim 1, The analysis process is a fusion analysis process between patent classification symbols, The convergence analysis process between the patent classification symbols may be performed by analyzing patent documents including two or more patent classification symbols on the same patent classification system among the patent documents included in the set of patent documents to be processed, Performing association analysis, frequent pattern analysis, correlation analysis, and clustering between two or more patent categories, In the fusion analysis processing between the patent classification symbols, When two or more patent classification symbols are included in one patent document in the same patent classification system, if the two or more patent classification symbols are not the highest patent classification symbol in the patent classification system, , The fusion analysis processing is performed for each of the two or more patent classification symbols including the upper patent classification symbol up to a predetermined level of the two or more patent classification symbols Information fusion method. The method according to claim 1, The convergence analysis processing may be performed for individual patent documents constituting the set of patent documents to be processed by 1) fusion analysis processing between at least two technical keywords generated based on text information extracted from at least one field of one patent document , 2) the analysis of the convergence between the co-inventors on the patent document with the co-inventor, 3) the analysis of the convergence between the co-applicants on the patent document with the co-applicant, and 4) the citation included in the citation information of one patent document And a fusion analysis process between patent documents, The fusion analysis processing may include at least one of association analysis, frequent pattern analysis, correlation analysis, and clustering between the two or more technical keywords, the co-inventor, the co-applicant, and the joint patent documents included in the patent documents The patent information fusion analysis method comprising the steps of: delete The method according to claim 1, The information about the line connecting the node and the node further includes numerical information, Receiving the numerical value or selection information for the line from the user computer; Extracting, from the patent information database including the set of patent documents to be processed, the found patent document information corresponding to the numerical information; And And transmitting the extracted patent document information to the user computer. A system for providing patent convergence analysis information to a user's computer, The system Document mast DB; Patent classification code mast DB; A patent analysis intelligence module that performs at least one patent analysis process on the specified set of analyzed patent documents; A sub-patent classification code preprocessing module unit, The sub-patent classification code preprocessing module may, when the first patent classification code is given, if the first patent classification code corresponds to a predetermined level on the patent classification code system to which the first patent classification code belongs, The information is processed so as to include the patent information corresponding to the sub-patent classification code of the classification symbol, And a patent analysis intelligence module for performing at least one patent analysis process on the specified set of analyzed patent documents, When the patent information analysis intelligence module generates the analysis information, the analysis information includes (1) a method of processing all the patent classification including patent classification symbols included in the processing target document set, and (2) main patent classification). In case of processing including all of the above patent classification, a) a method of treating all patent classes with equal weight, and b) a method of processing main patent classification (main patent classification) patent classification and sub-patent classification are applied with different weights to generate analytical information, A method of receiving the set of patent documents to be analyzed is selected from among 1) a method of receiving input from a user computer as a search formula, 2) information constituting a bibliography included in the predetermined patent document presented from the user computer to the user computer A method for receiving selection information on at least one set of patent documents selected from at least one set of patent documents previously generated by the user and 4) At least one set of at least one set of conditions is provided, The patent information generated by the patent analysis intelligence module is at least one or more multidimensional analysis information generated through at least one predetermined analysis process on the specified set of patent documents to be processed, The individual dimensions constituting the multidimensional are selected from among patent documents contained in 1) time, 2) patent classification symbol, 3) applicant, 4) inventor, 5) agent, 6) country, 7) Any one or more, The predetermined analysis process is a fusion pattern analysis process, Wherein the multidimensional analysis information includes at least one predetermined counting value for the at least one individual dimension, a calculation value by at least one predetermined calculation formula, a query by at least one quality consciousness, A result value, and a processing result value for at least one or more analysis processing algorithms, The patent analysis intelligence module includes a function of visualizing the fusion analysis processing result as a network diagram including at least two nodes, Wherein the visualization process further includes at least one of information about the node and information about a line connecting the node and the node. The method according to claim 6, The analysis process is a fusion analysis process between patent classification symbols, The convergence analysis process between the patent classification symbols may be performed by analyzing patent documents including two or more patent classification symbols on the same patent classification system among the patent documents included in the set of patent documents to be processed, Performing association analysis, frequent pattern analysis, correlation analysis, and clustering between two or more patent categories, In the fusion analysis processing between the patent classification symbols, When two or more patent classification symbols are included in one patent document in the same patent classification system, if the two or more patent classification symbols are not the highest patent classification symbol in the patent classification system, , The fusion analysis processing is performed for each of the two or more patent classification symbols including the upper patent classification symbol up to a predetermined level of the two or more patent classification symbols. Fusion analysis delivery system. The method according to claim 6, The convergence analysis processing may be performed for individual patent documents constituting the set of patent documents to be processed by 1) fusion analysis processing between at least two technical keywords generated based on text information extracted from at least one field of one patent document , 2) the analysis of the convergence between the co-inventors on the patent document with the co-inventor, 3) the analysis of the convergence between the co-applicants on the patent document with the co-applicant, and 4) the citation included in the citation information of one patent document And a fusion analysis process between patent documents, The fusion analysis processing may include at least one of association analysis, frequent pattern analysis, correlation analysis, and clustering between the two or more technical keywords, the co-inventor, the co-applicant, and the joint patent documents included in the patent documents The patent information fusion analysis providing system.

Images