JP4505236B2 - Patent / utility model literature search system and program - Google Patents

Description

  The present invention relates to a system for searching for documents by giving text (character string) as a keyword, and in particular, a system for efficiently searching for documents such as patents and utility model gazettes, and a program for realizing the system with a computer About.

  2. Description of the Related Art Conventionally, a document search system that extracts and presents a document including a character string by giving a text (character string) as a keyword to a database storing document information has been widely used in various fields. .

  Further, as a type of the above-described conventional document retrieval system, a system that retrieves patent / utility model publications using keywords is also known. For example, if you enter “technical terms”, “applicant”, and “inventor (inventor)” as keywords on the JPO website for beginners who are not so familiar with patent information, the corresponding patent / utility model bulletin A service for displaying information is provided (see Non-Patent Document 1). In addition to the JPO, search services for documents related to patents and utility models have been provided by various vendors.

Furthermore, a so-called publication distribution system is also known in which only publications that meet the conditions specified by the user are extracted from a patent information database such as a patent office and delivered to the user. If a patent document distribution system is used, for example, each time a patent document in a specific technical field is issued, or each time a patent document including a specific keyword is issued, the patent document is distributed. It is possible to obtain the patent documents of interest without taking the time and effort of searching.
“Patent Electronic Library of Japan Patent <Search for Beginners>” [online], [Search February 9, 2004], Internet <URL: http // www2.ipdl.jpo.go.jp / BE0 / index.html>

  By the way, in Japan, a huge number of patent / utility model publications are issued every year, so it may be very difficult to narrow down the number of documents hit by keyword search. Conventional search services or publication distribution systems generally use complex search expressions that combine multiple keywords with AND or OR conditions, or specify conditions other than keywords (for example, publication date) By doing so, an attempt is made to efficiently narrow down the documents required by the searcher.

  However, even if the search results are narrowed down by such a conventional method, the searcher eventually needs to carefully read the contents of the gazette and determine whether or not the search results are appropriate.

  An object of the present invention is to provide a patent / utility model literature search system that can more efficiently extract patent documents required by a searcher.

  In order to achieve the above object, a first patent / utility model literature search system according to the present invention corresponds to a high-order concept keyword corresponding to a constituent requirement of the invention to be searched and a subordinate concept of the constituent requirement. Based on the standard description format of the claim from the input section that allows the user to input the subordinate concept keyword and the text information of the claim of the patent or utility model document, the constituent requirements of the invention described in the claim are satisfied. A superordinate concept keyword extracting unit that extracts a superordinate conceptual keyword, and the superordinate concept keyword extracting unit based on a standard description format of the part from text information of a part other than the claims of the patent or utility model document A subordinate concept keyword extracting unit for extracting a subordinate concept keyword corresponding to a subordinate concept of the superordinate concept keyword extracted by A matching rate is calculated by comparing the set of higher concept keywords and lower concept keywords received with the set of higher concept keywords and lower concept keywords extracted by the upper concept keyword extraction unit and lower concept keyword extraction unit. Based on the rate calculation unit, the match rate calculated by the match rate calculation unit, a search result determination unit that selects a patent or utility model document to be presented to the user as a search result, and the search result determination unit And a display unit for displaying a patent or utility model document as a search result.

  In the first patent / utility model literature search system, when the subordinate concept keyword extraction unit extracts a subordinate concept keyword, based on a standard description format of a part other than the claim of the patent / utility model literature. The lower concept keyword according to a more preferable aspect is identified, and the matching rate calculation unit calculates a matching rate by assigning higher weight to the lower concept keyword according to a more preferable mode than other lower concept keywords. It is preferable.

  In order to achieve the above object, a second patent / utility model literature search system according to the present invention provides a patent or utility model literature extracted from a patent / utility model literature database by a distribution server based on a predetermined condition. And a publication number designating section for extracting all or part of the publication numbers of patents or utility model documents accumulated from the document storage section, and extracted by the publication number designating section. A publication number is sent to the distribution server, and a patent or utility model document including the publication number in the specification is extracted from the patent / utility model document database and distributed.

  In order to achieve the above object, a third patent / utility model literature search system according to the present invention uses a patent or utility model literature extracted from a patent / utility model literature database by a delivery server based on delivery conditions. A document storage unit for storing, a distribution condition storage unit for storing information on the distribution conditions, an input unit for inputting text data of application documents, the distribution condition storage unit, and a search keyword from the distribution conditions Based on the search keyword extraction unit to be extracted and the search keyword extracted by the search keyword extraction unit, the text data of the application document input from the input unit is searched, and the appearance status of the search keyword in the text data For each search keyword, and the appearance status obtained by the appearance status judgment unit is a predetermined condition. For a search keyword that satisfies the criteria, a document extraction unit that extracts a patent document including the search keyword from the document storage unit, and a display unit that displays an extraction result by the document extraction unit in a manner according to the appearance status of each search keyword It is characterized by comprising.

  In order to achieve the above object, a fourth patent / utility model literature search system according to the present invention provides a patent or utility model literature extracted from a patent / utility model literature database by a distribution server based on predetermined conditions. A document storage unit that stores information, an incidental information storage unit that stores additional information with respect to the information described in the document, and a display unit that displays the contents of patents or utility model documents stored in the document storage unit And, when displaying the contents of a patent or utility model document on the display unit, an additional information search unit that extracts additional information for the information described in the document from the incidental information storage unit, Information extracted by the supplementary information search unit is displayed on the display unit together with the document.

  In order to achieve the above object, a first program according to the present invention provides a user with a high-level concept keyword corresponding to a component requirement of the invention to be searched and a low-level concept keyword corresponding to a subordinate concept of the component requirement. From the input step to be inputted to the text and the text information of the claim of the patent or utility model document, based on the standard description format of the claim, the superordinate conceptual keyword corresponding to the constituent requirements of the invention described in the claim A superordinate concept keyword extraction step and a superordinate concept extracted by the superordinate concept keyword extraction unit from text information of a part other than the claims of the patent or utility model document based on a standard description format of the part A subordinate concept keyword extracting step of extracting a subordinate concept keyword corresponding to a subordinate concept of the keyword; and the input step A match is calculated by comparing the set of upper concept keywords and lower concept keywords inputted with the set of upper concept keywords and lower concept keywords extracted in the upper concept keyword extraction step and the lower concept keyword extraction step. Selected by the rate calculation step, the search result determination step for selecting a patent or utility model document to be presented to the user as a search result based on the match rate calculated by the match rate calculation step, and the search result determination step A command for causing the computer to execute a patent / utility model literature search process including a display step of displaying the patent or utility model literature as a search result.

  In the first program, when a lower concept keyword is extracted in the lower concept keyword extraction step, a more preferable mode is applied based on a fixed description format of a part other than the claims of the patent / utility model literature. It is preferable that lower concept keywords are identified, and in the matching rate calculation step, the matching rate is calculated by assigning higher weights to lower concept keywords according to a more preferable aspect than other lower concept keywords.

  In order to achieve the above object, the second program according to the present invention stores, in a document storage unit, a patent or utility model document extracted from a patent / utility model document database based on a predetermined condition by a distribution server. A publication accumulation step, a publication number extraction step for extracting all or part of publication numbers of patents or utility model documents accumulated from the literature accumulation section, and a publication number extracted by the publication number extraction step. A patent / utility model document search process including a search / distribution step for extracting and distributing a patent or utility model document including the publication number in the specification to the distribution server and extracting the patent or utility model document from the patent / utility model document database. Contains instructions to be executed.

  In order to achieve the above object, the third program according to the present invention stores a patent or utility model document extracted from a patent / utility model document database based on distribution conditions by a distribution server in a document storage unit. A document accumulation step, an input step for inputting text data of an application document, a search keyword extraction step for extracting a search keyword from the distribution conditions with reference to a distribution condition storage unit storing information on the distribution conditions, Based on the search keyword extracted in the search keyword extraction step, the text data of the application document input in the input step is searched, and the appearance status for each search keyword is obtained for the search keyword appearance status in the text data Determined by the determination step and the appearance status determination step A document extraction step for extracting a patent document including the search keyword from the document storage unit for a search keyword that satisfies a predetermined condition for the current situation, and an extraction result of the document extraction step in accordance with an appearance state of each search keyword And a command for causing the computer to execute a patent / utility model literature search process including a display step.

  In order to achieve the above object, the fourth program according to the present invention stores a patent or utility model document extracted from a patent / utility model document database by a distribution server based on a predetermined condition in a document storage unit. And ancillary information for extracting additional information for the information described in the document to be displayed, which is a patent or utility model document stored in the document storage unit. And a command for causing the computer to execute a patent / utility model document search process including a search step and a display step of displaying the information extracted in the incidental information search step together with the patent or utility model document on a display unit.

  According to the present invention, it is possible to provide a patent / utility model literature search system that can more efficiently extract patent documents required by a searcher.

Hereinafter, specific embodiments of the present invention will be described with reference to the drawings. In the following, documents to be searched by the search system according to the present invention may be collectively referred to as “patent documents”, but this is not intended to exclude documents related to utility models, but documents for both patents and utility models. including. In addition to patent publications under the current law, patent publications, patent publications, utility model registration bulletins, etc., patent documents include patent publications, utility model publications, utility model publications under the old law, etc. Is also included.
(Embodiment 1)
A patent / utility model information retrieval system according to an embodiment of the present invention will be described below. FIG. 1 is a block diagram showing a schematic configuration of a patent / utility model information retrieval system 100 according to the present embodiment.

  The patent / utility model information search system 100 of the present embodiment can perform a more precise search according to the conditions given by the user from the patent documents that have been narrowed down to some extent by conventional keyword search and other condition designations. System. Therefore, the patent / utility model information search system 100 includes at least an input unit 1, a patent document storage database 2, a search execution unit 3, and a display unit 4, as shown in FIG.

  The input unit 1 is a configuration for allowing a user to input a search condition, and the hardware configuration thereof is arbitrary on condition that text information can be input. The patent document accumulation database 2 includes only patent documents extracted from the patent / utility model bulletin database 5 held by the JPO or the patent information provider as those meeting a certain condition. Accumulated. Extraction of patent documents from the patent / utility model bulletin database 5 is executed by the distribution server 6. That is, the user designates the conditions of the required patent document from the input unit 1 of the patent / utility model information search system 100 to the distribution server 6 in advance. As a result, each time a document that matches the condition is added to the patent / utility model bulletin database 5, the distribution server 6 extracts the document and distributes it to the patent / utility model information search system 100. For example, if a user designates codes (international patent classification, F-term, etc.) in a technical field to be investigated, patent documents in the technical field can be collected and accumulated. Further, if a specific applicant name or identification number is designated as a condition relating to the applicant, for example, patent documents relating to applications of competitors can be collected and stored. Alternatively, if keywords such as technical terms are designated, patent documents including specific keywords can be collected.

  The patent document accumulation database 2 can be realized by a storage device such as a hard disk, but can also be realized as an independent server device or the like. As will be described in detail later, the patent documents stored in the patent document storage database 2 must have a file format that can be referred to as text information. In addition, between the patent / utility model information retrieval system 100 and the distribution server 6 and between the distribution server 6 and the patent / utility model gazette database 5, a wide area network such as the Internet or a dedicated line, if necessary. Connected through.

  The search execution unit 3 performs a more detailed search on the documents stored in the patent document storage database 2 according to the search condition input by the user from the input unit 1, and outputs the result to the display unit 4. The search execution unit 3 is realized by peripheral devices such as a CPU and a memory of a computer. The display unit 4 is a display for displaying the search result obtained by the search execution unit 3, and its hardware configuration is arbitrary on the condition that at least text information can be displayed. In addition, the input unit 1 and the display unit 4 can be configured as one case of hardware (for example, a touch panel display).

  The operation of the patent / utility model information search system according to this embodiment according to the above-described configuration will be described below with reference to FIG.

  In this patent / utility model information retrieval system, retrieval utilizing the subordinate concept of the invention is possible. Here, for example, a patent document related to the invention of “X consisting of A, B, and C” will be described on the assumption that a search is performed using subordinate concepts of A, B, and C. A, B, and C are constituent elements of the invention, and are expressed in a high-level conceptual expression. X corresponds to the name of the invention. In general, when the invention is described in the scope of claims or utility model registration claims, "A that is ... and B that contains ... Each component requirement is given a modifier such as “X consisting of C,” but “A”, “B”, and “C” used as keywords below are noun parts excluding the modifier. .

First, the user inputs texts of three keywords “A”, “B”, and “C” (hereinafter, referred to as higher concept keywords) from the input unit 1 (step S1 in FIG. 2). Next, the user arbitrarily inputs keywords A ₁ , A ₂ ,... Which are subordinate concepts of “A” from the input unit 1. Similarly, keywords B ₁ , B ₂ ,..., C ₁ , C ₂ ,... That are subordinate concepts of “B” and “C” are also arbitrarily input (step S2). The number of keywords input as a subordinate concept is arbitrary, and it does not necessarily have to be plural. For example, when X is “dental material” and “A” is “monomer”, the subordinate keywords A ₁ , A ₂ ,..., For example, “acryl”, “urethane”, “ For example, “vinyl alcohol” may be input.

Upper conceptual keywords A, B, C and their lower conceptual keywords A ₁ , A ₂ ,..., B ₁ , B ₂ ,..., C ₁ , C ₂ ,. Is sent to the search execution unit 3. After the initial value of i is set to 1 in step S3, the search execution unit 3 performs the following steps S4 to S6 on the patent document i stored in the patent document storage database 2. Note that i is a number uniquely assigned to each patent document stored in the patent document storage database 2.

  First, the search execution part 3 inputs the text information of the patent document i, and extracts the high-order concept keyword showing a structural requirement from the text of a claim (step S4). In the patent literature, a claim number is assigned to each claim using a delimiter. Thereby, the search execution part 3 can recognize easily which part in text information corresponds to description of a claim. Further, in general, the description of a claim is typical, and includes “Y composed of α, β, γ”, “Y including α, β, and γ”, “α, β, γ”. It is often described as “Y”. Therefore, the search execution unit 3 refers to the text information of the patent document i, and uses this regularity, for example, a noun immediately before the punctuation mark, a noun before the phrase “to”, or “ ”,“ Consisting of ”,“ comprising ”,“ comprising ”,“ consisting of ”,“ comprising ”, etc. It is possible to extract high-order concept keywords α, β, γ,. Note that such extraction of nouns from text data can be performed using a known method such as morphological analysis.

Next, the search execution unit 3 extracts lower concept keywords described in the patent document for each of the higher concept keywords α, β, γ,... Extracted in step S4 (step S5). . In the patent literature, for example, “α is α ₁ , α ₂ , α ₃ , or α ₄ ” or “as α, α ₁ , α ₂ , α ₃ , Or a regular description such as “α ₄ is used” is often used. Accordingly, the search execution unit 3 extracts nouns from the sentence including the higher concept keyword (for example, α), thereby extracting the lower concept keywords α ₁ , α ₂ , α ₃ , α ₄ from the above example sentence, for example. Can do. Alternatively, “α is preferably α ₁ , α ₂ , α ₃ , or α _4. ” Or “α is preferably α ₁ , α ₂ , α ₃ , or α ₄ . Such expressions are also frequently used in the specification. Therefore, extracting a noun from a sentence including the letters “good” together with the higher concept keyword “α” is also effective in appropriately extracting the lower concept keywords.

In addition, descriptions such as “α is α ₁ , α ₂ , α ₃ , or α ₄ , preferably α ₁ or α ₂ ” are often used in the specification. In this _{case, α 1, α 2, α} 3, of alpha _{4, 1} alpha mentioned as a preferred embodiment, the alpha ₂ is other subgeneric alpha _3, may be higher than alpha ₄ weighting.

Next, the search execution unit 3 obtains a set of upper concept keywords and lower concept keywords input by the user in step S2, and a set of upper concept keywords and lower concept keywords extracted from Patent Document i in step S4, respectively. By comparison, the coincidence rate of patent document i with respect to the condition input by the user is calculated (step S6). For example, the set of upper concept keywords input by the user in step S1 is {A, B, C}, and the set of lower concept keywords input by the user in step S2 is {A ₁ , A ₂ , A ₃ , B ₁ , B ₂ , C ₁ , C ₂ , C ₃ , C ₄ }, and the set of superordinate concepts extracted from the patent document in step S4 is {A, B, D, E}, and the subordinate concepts extracted in step S5 If the set is {A ₁ , A ₄ , A ₅ , B ₁ , B ₃ , C ₂ , C ₃ }, the matching elements in the set of superordinate concept keywords are {A, B} The matching elements in the set of subordinate concept keywords are {A ₁ , B ₁ , C ₂ , C ₃ }. The search execution unit 3 determines a match rate according to the number of matching elements for each of the superordinate concept and the subordinate concept as a numerical value based on a predetermined mathematical expression.

  For example, when the number of matches of the superordinate concept keyword is n and the number of matches of the subordinate concept keyword is m, p is the score when one superordinate concept keyword is matched, and q is the score when one subordinate concept keyword is matched. The value obtained from n × p + m × q can be considered as the coincidence rate. However, this calculation formula is merely an example, and any calculation formula can be used. In addition, as described above, when higher weights are given to subordinate concepts listed as particularly preferable aspects in the specification than other subordinate concepts, the matching rate should be calculated in consideration of the weights. May be.

  Through the processes in S4 to S6 described above, the coincidence rate of Patent Document i with respect to the conditions input by the user is calculated. The matching rate calculated for Patent Document i is stored in a memory (not shown) in the search execution unit 3. Then, in step S7, it is determined whether or not the calculation of the coincidence rate has been completed for all patent documents stored in the patent document storage database 2, and if not completed (the result of step S7 is NO), step In S8, the value of i is incremented by 1, and the process returns to Step S4. The above processing is repeated until the calculation of the coincidence rate is completed for all patent documents stored in the patent document storage database 2.

  When the calculation of the coincidence rate is completed for all patent documents accumulated in the patent document accumulation database 2 (the result of step S7 is YES), the search execution unit 3 is based on the coincidence rate stored in the memory. Then, out of all patent documents stored in the patent document storage database 2, a document having a matching rate higher than a predetermined value, or a document having a predetermined number of documents in descending order of matching rate, is extracted and displayed. 4 (step S9).

  In the present embodiment, the lower concept keyword is extracted from the constituent elements of the invention to be searched (upper concept keyword) and the matching rate is calculated. However, the claims of the patent document (the scope of the utility model registration request) The keywords automatically extracted from the parts other than () are not limited to keywords representing the subordinate concepts of the constituent elements of the claims. For example, when the constituent requirement keywords extracted from the claims are A, B, and C, when nouns are extracted from sentences including any of “A”, “B”, and “C”, A, B, In some cases, a keyword representing a higher-order concept than C and a keyword representing a concept parallel to A, B, and C may be extracted. Therefore, as a keyword to be input in step S2 of FIG. 2, a keyword of a higher concept than the keyword input in step S1 or a keyword of a concept parallel to the keyword input in step S1 is input instead of the lower concept keyword. By doing so, it is possible to search for patent documents including keywords of a higher concept or parallel concepts than the constituent requirements of the invention to be searched.

  Further, as a further modification, a keyword representing the content of the component component or the like may be extracted. For example, in order to search for patent documents related to the invention such as “10 to 20% by weight of A, 20 to 30% by weight of B,... In patent literature, a sentence including the keyword “A” and the keyword “% by weight” is searched, and the numerical value immediately before the phrase “% by weight”, or this numerical value and “˜”, “to”, or A numerical range represented by a numerical value connected by a standard word “or” is extracted. Then, the coincidence ratio may be calculated based on the degree of overlap between the extracted numerical value or numerical value range and the content (10 to 20% by weight for A in the above example) input by the user. This modification has an advantage that an effective search can be made especially for the invention in the chemical field.

  As described above, according to the patent / utility model information search system according to the present embodiment, not only a simple search using a keyword but also a configuration given by a keyword based on expressions that are used in a patent document. It is possible to search the subordinate concept and content of the requirement to calculate the coincidence rate, and to present a patent document having a high coincidence rate to the user. As a result, it is possible to obtain a highly accurate search result reflecting the substantial similarity of the invention rather than a simple search using keywords.

In the present embodiment, an example is shown in which a search is performed using all of A, B, and C, which are constituent requirements of the inventive concept X to be searched, as higher-level concept keywords. However, it is not necessary to perform a search for all the configuration requirements of the inventive concept. For example, only a new configuration requirement or a characteristic configuration requirement among all the configuration requirements of the inventive concept may be used as a higher concept keyword.
(Embodiment 2)
A patent / utility model information search system according to another embodiment of the present invention will be described below. FIG. 3 is a block diagram showing a schematic configuration of the patent / utility model information search system 200 according to the present embodiment. In addition, about the structure which has the function similar to the structure demonstrated in Embodiment 1, the same referential mark as Embodiment 1 is attached, and detailed description is abbreviate | omitted. The same applies to other embodiments described later.

  As shown in FIG. 3, the patent / utility model information search system 200 includes a publication number designation unit 11 in addition to at least the patent document accumulation database 2 and the display unit 4.

  The patent document storage database 2 stores the patent documents distributed by the distribution server 6 as described in the first embodiment. The distribution server 6 extracts from the patent / utility model bulletin database 5 patent documents that meet the conditions specified from the input unit 1 of the patent / utility model information retrieval system 200 and distributes them to the patent / utility model information retrieval system 200. To do. For example, if codes (technical patent classification, F-term, etc.) of a technical field to be searched are specified from the patent / utility model information retrieval system 200, patent documents in the technical field are collected and accumulated. Can do. Further, if a specific applicant name or identification number is designated as a condition relating to the applicant, for example, patent documents relating to applications of competitors can be collected and stored. Alternatively, if a keyword such as a technical term is specified, patent documents including the keyword can be collected.

  The patent / utility model information search system 200 of the present embodiment gives a publication number of a patent document from the publication number designation unit 11 to the distribution server 6, and there is no patent document including the publication number in the specification in the distribution server 6. It further has a function to search. In the specifications of patent applications, publication numbers of patent documents are often described as prior art related to the application in the description of the prior art. In particular, in applications after the 2002 Amendment Act is enacted, it is necessary in principle to describe prior art document information in the specification in accordance with Article 36, Paragraph 4, Paragraph 2 of the Patent Act. Therefore, by searching for a patent document that includes the designated publication number in the specification, it is possible to efficiently find the related subsequent application patent document.

  Note that, for example, the publication method of the publication number may be different even in the same publication, as shown in, for example, Japanese Patent Publication No. 12345, Japanese Laid-Open Patent Publication No. 45-12345, and Japanese Laid-Open Patent Publication No. 45-012345. Therefore, it is also preferable to use an ambiguous search when searching for publication numbers. Further, for example, when both a public gazette and a patent gazette are issued, a search method that recognizes them as the same patent document may be used.

  Further, as a publication number to be searched, a number directly input by the user from the input unit 1 may be sent to the publication number designation unit 11. Alternatively, the publication number designation unit 11 may extract the publication number from all or part of the patent documents stored in the patent document storage database 2. In the former case, that is, when the user directly inputs the publication number, it is possible to search related patent application documents for any patent document desired by the user. On the other hand, in the latter case, it is possible to search for subsequent patent documents related to the patent documents stored in the patent document storage database 2. In addition, among the patent documents stored in the patent document storage database 2, the publication number designation unit 11 may extract the publication number according to the distribution conditions and the like. An example of the method will be described below.

  FIG. 4 shows an example of patent document information stored in the patent document storage database 2. In the example of FIG. 4, in the patent document storage database 2, for each stored patent document, an item number (serial number in the database), a distribution condition that the patent document matches, a publication number, and the patent document A link to a file (text or pdf) is stored. The patent documents of item numbers 1 and 4 are patent documents distributed as a result of matching the condition designation regarding the applicant. The patent documents of item numbers 2 and 3 are patent documents distributed when the condition specification related to the keyword (including the keyword “dental material”) is matched. The storage form of the patent document information in the patent document storage database 2 is not limited to this example. For example, in addition to the publication number, additional information such as the publication date and the applicant is also extracted from the patent document. It may be held in the same manner as the publication number.

  Then, for example, when the user instructs the input unit 1 to search for a patent application document related to a patent document that matches the condition specification related to the keyword “dental material”, the publication number specification unit 11 displays the items shown in FIG. The publication numbers of numbers 2, 3,... Are extracted and sent to the distribution server 6. Then, the distribution server 6 searches the patent / utility model bulletin database 5 for a patent document including the publication number sent from the publication number designation unit 11 in the specification. The distribution server 6 distributes the search result to the patent / utility model information search system 200.

As a result of the search for the patent application patent documents as described above, for example, it may be found that a patent document is described as a prior art document in patent documents in completely different technical fields. Accordingly, the patent / utility model information retrieval system 200 according to the present embodiment may be particularly effective in, for example, retrieval of a material use invention.
(Embodiment 3)
A patent / utility model information search system according to still another embodiment of the present invention will be described below. FIG. 5 is a block diagram showing a schematic configuration of the patent / utility model information retrieval system 300 according to the present embodiment.

  The patent / utility model information search system 300 according to the present embodiment creates a set of related patent documents when the user is creating application documents in order to efficiently find patent documents related to the user's application. And has a display function.

  Therefore, the patent / utility model information search system 300 includes at least an input unit 1, a patent document storage database 2, a search execution unit 21, a distribution condition storage unit 22, and a display unit 4, as shown in FIG. Yes.

  The user inputs text data of application documents (including application, claims (utility model registration request), description, abstract) from the input unit 1. The input from the input unit 1 may be sequential text input using a keyboard or the like, or may be data input by reading text data of a completed application document from a storage medium or a communication line. .

  Also in the present embodiment, the distribution server 6 extracts patent documents that match the distribution conditions specified by the user from the input unit 1 from the patent / utility model bulletin database 5 as in the above-described embodiments. Distribute to the utility model information retrieval system 300. The patent document storage database 2 stores patent documents distributed from the distribution server 6. The distribution conditions specified by the user from the input unit 1 are sent to the distribution server 6 and stored in the distribution condition storage unit 22. FIG. 6 shows an example of distribution conditions stored in the distribution condition storage unit 22. In the example of FIG. 6, as delivery conditions specified by the user, (1) the applicant is company A, (2) the inventor is △ Yamao, (3) “dental material” and “monomer” (4) Applicant is Company B and the inventor is △ kawa ○ child, (5) Applicant is Company C and includes the keyword “ceramic” Such distribution conditions are stored in the distribution condition storage unit 22. Note that the contents and description method of the distribution conditions are merely examples, and arbitrary distribution conditions can be designated.

  FIG. 7 is a flowchart showing the operation of the search execution unit 21 according to the present embodiment. The search execution unit 21 first refers to the distribution condition storage unit 22 and extracts keywords stored as distribution conditions (step S21). For example, in the case of FIG. 6, at least the keywords “dental material”, “monomer”, and “ceramic” are extracted.

  Next, the search execution unit 21 searches whether or not the keywords extracted as described above exist in the text data of the application document input from the input unit 1, and these keywords exist in the application document. If so, the number of appearances of each keyword is counted (step S22). When the keyword search is completed for the entire application document, the search execution unit 21 determines which keyword has the highest number of appearances (step S23).

  Next, the search execution unit 21 accesses the patent document accumulation database 2 and extracts the patent documents including the keyword having the highest appearance frequency (step S24). Then, the extracted patent document is displayed on the display unit 4 (step S25).

  In the above description, in step S24, only patent documents including the keyword having the highest number of appearances are extracted. However, a similar search is performed for a plurality of keywords having a relatively high number of appearances, and each keyword is searched. In step S25, the display mode may be made different so that sets of patent documents obtained as search results can be distinguished from each other. For example, in the example of FIG. 6, if the keyword having the highest number of appearances in the application document is “monomer” and the next most frequently occurring keyword is “ceramic”, a list of patent documents including “monomer” It is conceivable to display a list of patent documents including “Ceramic” with priority and color-coded display. Further, instead of the number of appearances, a keyword used as a search keyword may be determined based on the appearance frequency or the like.

Through the above processing, the patent / utility model information retrieval system 300 efficiently retrieves patent documents presumed to be highly relevant to the application documents being prepared by the user from the patent documents stored in the patent document storage database 2. Can be extracted and presented to the user. This is effective for user application support.
(Embodiment 4)
A patent / utility model information search system according to still another embodiment of the present invention will be described below. FIG. 8 is a block diagram showing a schematic configuration of the patent / utility model information search system 400 according to the present embodiment.

  The patent / utility model information search system 400 according to the present embodiment can display incidental information (described later) for each of the patent documents stored in the patent document storage database 2. Therefore, as shown in FIG. 8, the patent / utility model information search system 400 includes at least an input unit 1, a patent document storage database 2, and a display unit 4, an auxiliary information database 31 and an auxiliary information search unit 32. I have.

  The patent document storage database 2 stores the patent documents extracted and distributed from the patent / utility model gazette database 5 by the distribution server 6 as in the first embodiment. In the incidental information database 31, incidental information (hereinafter referred to as “accompanying information”) with respect to information described in the patent document is stored.

  Examples of incidental information include (1) information related to the inventor or applicant, (2) information related to prior art document information described in the specification, and the like. It is not limited and arbitrary information can be used.

  Further, as the information related to the inventor or applicant of the above (1), more specifically, the following examples can be considered, but the present invention is not limited to these, and any other information is used. Can do.

(1-1) Information suggesting the possibility that the inventor is a university teacher;
(1-2) If the applicant is a competitor of the user, information on the competitive technology field, competitive products, etc.
(1-3) Information indicating a transaction product, a transaction status, and the like when the applicant is a business partner of the user.

  For example, a case where the information (1-1) is used as the incidental information will be described. When the inventor is a university faculty member, there are three cases: the applicant is the inventor (teacher) himself, the research contract company, or the university. Therefore, it has been difficult to determine whether or not it is an invention of a university teacher only from the inventor and applicant fields.

  Therefore, in the patent / utility model information retrieval system 400, university teacher information including the name, address, university name, department name, career, etc. of the university teacher is stored in the incidental information database 31. Then, when displaying the patent documents stored in the patent document storage database 2 on the display unit 4, the incidental information search unit 32 uses the name of the inventor of the patent document as a keyword and the university faculty member of the incidental information database 31. Search for information. Then, if there is university faculty information that matches the inventor's name and address, the incidental information is displayed on the display unit 4 next to the inventor column of the patent document, for example, as shown in FIG. The contents of information stored in the database 31 (for example, address, university name, department name, career, etc.) are displayed as comments 51. As a result, the user can easily recognize that the inventor is a university professor without conducting his own research. If only the name of the inventor described in the patent document matches the university faculty member information, the content of the comment 51 may be a notation that suggests the possibility, for example, as shown in FIG. It ’s fine. Note that the content to be displayed as a comment and the display mode of the comment are not limited to these specific examples.

  Further, when the information (1-2) is used as the incidental information, the competitor information including the competition technical field and the competition product name is stored in the incidental information database 31 for one or more competitors. Keep it. When displaying the patent documents stored in the patent document storage database 2 on the display unit 4, the incidental information search unit 32 uses the applicant's name of the patent document as a keyword and displays competitor information in the incidental information database 31. Search for. If there is competitor information that matches the applicant's name, the content of the competitor information stored in the incidental information database 31 is displayed as a comment on the display unit 4 next to the applicant's column in the patent document. To do.

  In addition, when the information (1-3) is used as the incidental information, the incidental information database 31 stores customer information including transaction products and transaction statuses for one or a plurality of customers. . Then, when displaying the patent documents stored in the patent document storage database 2 on the display unit 4, the incidental information search unit 32 uses the applicant name of the patent document as a keyword and the supplier information in the incidental information database 31. Search for. If there is supplier information that matches the applicant name, the contents of the supplier information stored in the incidental information database 31 are displayed as comments on the display unit 4 next to the applicant column of the patent document. To do.

  Furthermore, when using the “information related to prior art document information described in the specification” in (2) above, each of the patent documents stored in the patent document storage database 2 is stored in the incidental information database 31. The bibliographic items of the prior art document and the prior art document related information including the link to the document are stored. When the patent documents stored in the patent document storage database 2 are displayed on the display unit 4, the incidental information search unit 32 uses the prior art document number included in the patent document as a keyword, Search related information related to prior art documents. If there is related art document related information that matches the prior art document number, the related art document related information stored in the incidental information database 31 is displayed on the display unit 4 next to the prior art document number in the specification. Display the contents of information as a comment.

  As described above, in the patent / utility model information search system 400 according to the present embodiment, the patent document stored in the patent document storage database 2 is described in the patent document when displayed on the display unit 4. Information related to the inventor, applicant, prior art document information, etc. is acquired from the incidental information database 31 and displayed. Thereby, when the user views the patent documents stored in the patent document storage database 2, the accompanying information can be easily viewed.

  In the above description, the patent / utility model information retrieval system 100, 200, 300, 400 implemented as an object (system) is illustrated as one aspect of the present invention. Is a program for realizing the above on a general-purpose computer. That is, the present invention can be implemented as a program for causing a computer to execute the processing described in each embodiment.

  INDUSTRIAL APPLICABILITY The present invention is useful as a patent / utility model information retrieval system capable of efficiently retrieving documents related to patents or utility models.

1 is a block diagram showing a schematic configuration of a patent / utility model information retrieval system according to an embodiment of the present invention. FIG. It is a flowchart which shows operation | movement of the patent and utility model information search system shown in FIG. It is a block diagram which shows schematic structure of the patent and utility model information search system concerning other embodiment of this invention. It is explanatory drawing which shows an example of the patent document information accumulate | stored in the patent document accumulation | storage database in the patent and utility model information search system shown in FIG. It is a block diagram which shows schematic structure of the patent and utility model information search system concerning further another embodiment of this invention. FIG. 6 is an explanatory diagram illustrating an example of distribution condition data stored in a distribution condition storage unit in the patent / utility model information search system illustrated in FIG. 5. It is a flowchart which shows operation | movement of the search execution part in the patent and utility model information search system shown in FIG. It is a block diagram which shows schematic structure of the patent and utility model information search system concerning further another embodiment of this invention. (A) And (b) is explanatory drawing which shows an example of the aspect by which the comment containing incidental information was displayed on the display part in the patent and utility model literature search system shown in FIG.

Explanation of symbols

100, 200, 300, 400 Patent / utility model information retrieval system 1 Input unit 2 Patent document accumulation database 3 Search execution unit 4 Display unit 5 Patent / utility model gazette database 6 Distribution server 11 Publication number designation unit 21 Search execution unit 31 Information database 33 Attached information search part 51 Comments

Claims (2)

An input unit that allows a user to input a superordinate conceptual keyword corresponding to a constituent requirement of the invention to be searched and a subordinate conceptual keyword equivalent to a subordinate concept of the constituent requirement;
From the text information of a claim in a patent or utility model document , a noun immediately preceding a fixed phrase of a sentence in a fixed description form of the claim is a superordinate concept corresponding to the constituent elements of the invention described in the claim a higher concept keyword extraction unit that extracts as a keyword,
The superordinate concept extracted from the text information of the part other than the claim of the patent or the utility model document by the superordinate concept keyword extraction unit for the noun in the sentence including the superordinate concept keyword in the standard description format of the part a lower concept keyword extracting unit that extracts a subordinate conceptual keywords corresponding to subordinate concept of keywords,
The matching rate by comparing the set of upper concept keywords and lower concept keywords input from the input unit with the set of upper concept keywords and lower concept keywords extracted by the upper concept keyword extraction unit and lower concept keyword extraction unit A match rate calculation unit for calculating
A search result determination unit that selects a patent or utility model document to be presented to the user as a search result based on the match rate calculated by the match rate calculation unit;
A display unit that displays the patent or utility model document selected by the search result determination unit as a search result ,
When the subordinate concept keyword extracting unit extracts subordinate concept keywords, the subordinate concept according to a more preferable aspect is selected from a sentence including the superordinate concept keywords in a fixed description format other than the claims of the patent / utility model literature. Identify conceptual keywords,
The coincidence rate calculating unit is more preferable for the lower conceptual keywords related to embodiment, to calculate the matching rate by performing a higher weighting than other lower conceptual keywords, patent utility model document retrieval system. An input step for causing a user to input a high-order concept keyword corresponding to a constituent requirement of the invention to be searched and a low-order concept keyword corresponding to a subordinate concept of the constituent requirement;
From the text information of a claim in a patent or utility model document, a noun immediately preceding a fixed phrase of a sentence in a fixed description form of the claim is a superordinate concept corresponding to the constituent elements of the invention described in the claim A high-level concept keyword extraction step to extract as keywords;
The superordinate concept extracted from the text information of the part other than the claim of the patent or the utility model document by the superordinate concept keyword extraction unit for the noun in the sentence including the superordinate concept keyword in the standard description format of the part A subordinate concept keyword extracting step of extracting as a subordinate concept keyword corresponding to the subordinate concept of the keyword;
The matching rate by comparing the set of higher concept keywords and lower concept keywords input in the input step with the set of higher concept keywords and lower concept keywords extracted in the upper concept keyword extraction step and lower concept keyword extraction step A match rate calculating step for calculating
A search result determining step for selecting a patent or utility model document to be presented to the user as a search result based on the match rate calculated by the match rate calculating step;
A display step of displaying the patent or utility model document selected by the search result determination step as a search result,
When extracting a lower concept keyword in the lower concept keyword extraction step, a subordinate according to a more preferable aspect from a sentence including the upper concept keyword in a standard description format of a part other than the claim of the patent / utility model document. Identify conceptual keywords,
In the matching rate calculation step, the lower concept keyword according to a more preferable aspect is weighted higher than other lower concept keywords to calculate the matching rate.

Images