JP2018085082A - Search method of patent document - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method enabling comprehensive search of a compliant document in a patent database even if there is no deep knowledge on patent classification of a specific technology.SOLUTION: A compliant document incremental search formula is created using a patent classification and a keyword of a master patent document set created by an arbitrary method, and a patent database search is performed using the search formula to create an OR set of a hit document set and the master patent document set. The OR set is screened to create a compliant document set, which is evaluated with a function of the compliant document set. As a result of a function evaluation, when it fails, documents in the master patent document set are increased by an arbitrary method and a compliant document search is performed continuously until the compliant document set will pass the function evaluation.SELECTED DRAWING: Figure 2

Description

The present invention relates to a patent document search method for creating a patent document set that comprehensively includes similar patent documents such as a document set used for creating a patent map.

A patent document that conforms to the search purpose (the patent document is hereinafter referred to as a conforming patent document or a conforming document) and that does not leak the search document (the patent document is referred to as a nonconforming patent document or a nonconforming document). In order to create a search expression that hits as little as possible, knowledge of patent classification in the field to be searched, keywords in the field to be searched, and knowledge of synonyms and similar words of the keyword are indispensable. It is very difficult to master the knowledge about the above patent classification and keywords in all the fields encountered, and even if there is some knowledge of the search target field, the relevant literature was collected comprehensively The fact is that there are many cases where I cannot be confident.

Based on such circumstances, it is composed of a method for comprehensively searching similar patent documents and a large number of patent documents without having deep knowledge about patent classification, keywords, synonyms and similar words of the keywords, etc. A method for efficiently retrieving relevant documents from a document set has been proposed.

As an example 1, Japanese Patent Application Laid-Open No. 2003-141164 automatically inputs other patent documents by using the F-term assigned to the patent document by simply inputting the number of the specific patent document without considering the patent classification or keyword in advance. A method has been proposed in which similarities are displayed with the number of common F-terms. However, all relevant documents such as those used for patent map creation are comprehensively collected and non-conforming documents are not included as much as possible. When trying to collect a document set, it is difficult to categorize relevant documents and non-conforming documents only by the number of patent classifications without using a search that uses keywords. It is difficult to find differences between these sets by comparing the number of patent classifications.

As an example 2, Japanese Patent Laid-Open No. 2003-223465 describes a patent document search system or a patent document search method that organizes synonym dictionaries indispensable in patent document search using natural language as a key and generates less leakage and noise. However, since this method needs to create a synonym dictionary for each field in units of IPC group or F-term theme fields, it is considered effective for searching patent documents in technical fields that appear repeatedly. In searching for patent documents in a field, creating a synonym dictionary for each field requires a wide knowledge about the technical field, so it is not always a patent document search method that can be immediately implemented.

JP2003-141164A JP2003-223465

Even if there is no deep knowledge about the patent classification of a specific technology, even if there is no deep knowledge about the synonyms and similar words of a specific technology, based on the patent classification and keywords of any master patent document set, the relevant documents in the patent database It is an object of the present invention to provide a method for exhaustively searching (hereinafter, this search is referred to as exhaustive relevant literature search).

In order to solve the above-described problems, the present invention employs means for solving the problems having the following characteristics.

The invention described in claim 1 is a patent document search method executed by a computer in which patent document analysis software and spreadsheet software are installed.
A matching document increasing search formula creating step for creating a matching document increasing search formula using the patent classification and keywords used in the master patent document set, and a patent database searching step for searching a patent database using the matching document increasing search formula; Screening a logical sum set of the document set hit in the patent database search step and the master patent document set to create an all-matched document set creating step and evaluating the all-matched document set by sensory evaluation If it passes, the above-mentioned all relevant document set is set as an exhaustive relevant document set,
If the all relevant document set is evaluated by sensory evaluation and rejected, the number of documents in the master patent document set is increased by an arbitrary method, and the relevant document increases until the all relevant document set passes by sensory evaluation. Performing a search formula creation step, the patent database search step, and the all relevant document set creation step;
The patent document search method is characterized in that the search for relevant documents is continued until the set of all relevant documents is passed through sensory evaluation.

In the invention according to claim 2, in the matching document increasing search formula creation step, a search query by hierarchy is created for every hierarchy of the keyword table by hierarchy of the master patent document set, and the created search queries by all hierarchies The logical product of all the hierarchical search expressions other than the hierarchical search expression is calculated for every hierarchical search expression for which it is desired to increase the number of documents in the document set including the matching reference that is not included in the logical product. Create a search formula that logically ORs all the created logical products, and use the search formula as a keyword for the relevant literature increase formula and the ratio of the number of documents by cumulative patent classification in the master patent document set as an evaluation scale The method for searching patent documents according to claim 1, wherein the patent classification of the corresponding literature increasing formula is used.

The invention described in claim 3 is a patent document search method that is executed by a computer in which patent document analysis software and spreadsheet software are installed.
A document set creation step of creating a document set X by randomly extracting the number of documents satisfying the keyword document number satisfaction degree reference value from the master patent document set,
An exhaustive relevant literature search formula creation step for creating an exhaustive relevant literature search formula using the patent classification used in the master patent literature set and a keyword extracted from the literature set X, and the exhaustive relevant literature search formula Patent document search using patent database search step, and performing comprehensive exhaustive document search by screening a logical sum set of a document set hit in the patent database search step and the master patent document set Search method.

In the invention according to claim 4, in the comprehensive relevant literature search formula, a hierarchical search formula is created for every hierarchy of the hierarchical keyword table of the patent document set X. Create a logical product of all the hierarchical search expressions other than the hierarchical search expression for every hierarchical search expression for every hierarchical search expression that wants to increase the number of documents in the document set including the hit documents that are not included in the logical product, Create a search formula that logically ORs all the created logical products, and determine the search formula as a keyword for the comprehensive relevant document search formula and the ratio of the number of documents by cumulative patent classification of the master patent document set as an evaluation scale The patent document search method according to claim 3, wherein the patent classification is used.

The invention according to claim 5 is created by randomly extracting the number of documents satisfying the keyword document number sufficiency standard value from the master patent document set, wherein the number of documents of the master patent document set is insufficient. If the document set X cannot be created, the number of documents in the master document set is increased by an arbitrary method until the number of documents satisfying the keyword document number satisfaction degree reference value is exceeded, and the increased number of relevant document sets is increased. The patent document search method according to claim 3, wherein an exhaustive relevant document is searched using the master patent document set as a master patent document set.

The search method of the patent document has the following effects.

Even those who are unfamiliar with patent classifications and synonyms of keywords of a specific technology can create a comprehensive reference document set.

Patent document analyzer Flow chart of a patent document search method for increasing the number of relevant documents using a master patent document set Master Patent Document Set a or c of Examples 1 and 3 Number of documents by cumulative patent classification of master patent document set a or c of Examples 1 and 3 Name and summary of invention of master patent document set a of Example 1 and number of documents by keyword in claims Hierarchical arrangement table of keywords extracted from the name and abstract of the invention of the master patent document set a of Example 1 and the claims Hierarchical keyword table of master patent document set a or c of Examples 1 and 3 Relevant literature increase search formula a or c of Examples 1 and 3 Retrieval formula for creating master document set b of embodiment 2 Scatter diagram for determining the document number satisfaction degree of the master patent document set b of Example 2 Number of documents by cumulative patent classification of master patent document set b of Example 2 Hierarchical Keyword List for Exhaustive Relevant Literature Search Formula b of Example 2 Extensive relevant literature search formula b of Example 2 Tolerance of the estimated value and the required number of sampled documents when the number of specific documents included in a specific patent document set is estimated by the sampling method Scatter chart for determining the number of documents in the master document set c in Example 3 Scatter chart for determining the document number satisfaction degree of the relevant document augmentation set c1 in Example 3 The number of documents by cumulative patent classification of the relevant document increase set c1 in Example 3 Hierarchical keyword table for an exhaustive relevant literature search formula created from the relevant literature augmentation set c1 of the third embodiment Comprehensive conforming literature search formula c of Example 3

Hereinafter, embodiments of a search method for patent documents of the present invention will be described with reference to the drawings. However, the embodiments of the present invention show the most preferable modes of the present invention, and the present invention is not limited thereto. .

As shown in FIG. 1, a computer in which patent document analysis software and spreadsheet software used for analyzing a patent document set, which is a collection of a plurality of patent documents to which patent classifications are assigned, such as a published patent gazette and a patent gazette, are installed. The structure of the patent document analyzer 1 provided with 2 is demonstrated.

This patent document analysis apparatus 1 has a computer 2 to which a control device 3, an arithmetic device 4, a storage device 5 and an I / O 6 are connected via a bus line, and on the input side of the I / O 6 of the computer 2 A keyboard 7, a mouse 8, an input device 9, a CD-ROM drive 10 including an MO drive, etc. are connected, and on the output side of the I / O 6, a display device 11 comprising a CRT (including a liquid crystal display), a printer 12 etc. are connected. Furthermore, it is possible to connect to an ASP patent database 13 such as SRPARTNER or NRI Cyber Patent Desk through the communication line 14. The patent database 13 stores bibliographic information and patent documents such as specification contents.

The main functions of the patent document analyzer 1 are as follows.
When the input device 9 is operated to open a file of a set of patent document to be analyzed using the patent document analysis software preinstalled in the storage device 5 of the computer 2, a front page is displayed. Then, select a data object such as search / list and proceed with the operation.

To create a patent map, a map is selected as a data object, a map creation condition setting screen is drawn on the display device 11, and various map creation conditions are set on the setting screen. Map (the first column is the number of documents to which each display item column such as patent classification, keyword, applicant, etc., the second column is assigned the keyword of the display item column, the first column is the second column A numerical table arranged in descending order using the number of documents as an alignment key is mapped, and switching between the numerical table and the map can be easily performed by the input device 9), or a reference number comparison map of two document sets by keyword (one column) The second column is the number of documents in the document set 1. The third column is the number of documents in the document set 2. The fourth column is the number of documents in the second column. The keyword array of the display items in the first column is a map of a numerical table that is arranged in descending order using the ratio in the fourth column as the alignment key. Can be easily made by the input device 9).

In addition, the numerical table of the various maps is copied and pasted to the sheet of the spreadsheet software displayed on a screen different from the numerical table screen, and the copied and pasted numerical table is used using the spreadsheet software. Addition or deletion of rows and columns, description of characters in each cell, deletion, arithmetic operation of numbers between cells, alignment of a specific range of data in ascending or descending order using characters in a specific row or column, etc. Processing can be done freely.

Retrieval based on various patent classifications and keywords for a patent document set stored in the storage device 5 as a search target, a logical sum set of files stored as objects of the patent document set stored in the storage device 5, and logic To create a difference set and logical product set, a search screen for patent information can be drawn on the screen of the display device 11 and can be freely searched and calculated using the input device 9 on the screen.

Searching for a set of patent documents stored in the storage device 5 as a search target, the search keyword is a keyword that partially matches a specific keyword of a specific patent document item, a keyword that matches the front, or a keyword that matches the back In the case, the specific keyword is input in the keyword field for narrowing down that is set in the keyword list screen of the specific patent document item that can be peeked from the patent information search screen, the narrowing type is partially matched, forward matching, When any desired type of backward matching is selected and narrowed down, a keyword list of specified conditions is drawn, so that a keyword to be used can be selected from the list.

A method for creating a collection of patent documents necessary for analysis is as follows. First, the input device 9 is operated to access the patent database 13 or the published patent publication CD set in the drive 10 such as a CD-ROM, Patent literature search is performed using a search expression composed of a patent classification, a keyword, etc., the hit literature set is named and saved in the storage device 5, and the patent literature saved in the storage device 5 The set is converted into a format that can be processed by the patent document analysis software using a data format conversion program installed in the storage device 5 in advance, and the converted patent document set is screened as necessary to analyze patent documents. A set is created and stored in the storage device 5.

Patent classification in this application means patent classification symbols such as IPC, FI section, class, subclass, main group, subgroup, and F-term theme code, theme code + viewpoint, F-term, etc. Means a document to which the patent classification is given.

The patent document items described in the present application are the generic name of the title of the invention, abstract, claims, full description, and combinations of these items.
An exhaustive relevant literature search formula is a search formula for exhaustively searching for patent literature that matches the search purpose.
An exhaustively relevant document set is a document set in which documents suitable for a search purpose are collected in an exhaustive manner.

A patent document search method for comprehensively collecting relevant documents using patent classifications and keywords of a master patent document set collected by an arbitrary method will be described with reference to FIGS. The explanation related to 2 is a case where a literature search is performed until the conformity document coverage of the augmented reference literature set obtained by screening passes the sensory evaluation. (Note 6) The case where the number of keyword saturation point compatible documents can be obtained, and the explanation related to claim 5 is the case where the number of keyword saturation point compatible documents in the master patent document set cannot be obtained.

The steps of the patent document search method relating to claims 1 and 2 are described with SA, the step relating to claims 3 and 4 with SB, and the step relating to claim 5 with SC.

1. A patent document search method SA1 and a master patent document set A related to claims 1 and 2 are created.

The creation method of the master patent document set A in this step may be any method. For example, a search formula created using keywords extracted from relevant patent documents, or keywords devised by the patent information surveyed by the patent information platform. There is a method in which a patent document search is performed and a hit document set is screened and created.

SA2, the patent classification to be used in the relevant document increase search formula A is determined using the (Note 2) cumulative patent classification document number ratio in the (Note 1) cumulative patent classification document count table of the master patent document set A as an evaluation scale.

(Note 1) The number of documents by cumulative patent classification of a specific relevant literature set (see, for example, FIG. 5, FIG. 11, and FIG. 17) is a search for increasing relevant literature from the patent classification used in the specific relevant literature set. Patent classification summary table for extracting patent classifications used in formulas and exhaustive relevant literature search formulas, and the patent classifications assigned to the documents in the relevant literature set are the documents to which the patent classifications are assigned The number of cases is arranged in descending order as the alignment key. (1) The patent classification column and the number of documents to which the patent classification of the patent classification column is assigned are described. (2) The document number column and the first number of the document number column. The patent classification is the first line, the logical sum of the 1st and 2nd patent classifications is the 2nd line, the logical sum of the 1st, 2nd and 3rd patent classifications is the 3rd line , Each row has a logical OR set of patent classifications up to the rank of each row (3) Cumulative patent classification string and search target as the specific relevant document set, search keyword as the cumulative patent classification of (3), and the number of documents hit by the search are described (4) The number of document hits by cumulative patent classification and the ratio of the number of documents in the (4) column to the total number of documents in the relevant document set are listed. (5) In the configured table, the creation method of the table is as follows: (1) Patent classification column (2) Document number column The number of documents classified by patent classification of the specific relevant document set created using the patent document analysis software The patent classification and the number of documents in the ranking table are described (for example, see columns (1) and (2) in FIG. 5, FIG. 11 and FIG. 17). Described using patent classification, (4) column by cumulative patent classification The number of documents in the contribution hit number column is obtained by performing a search on the patent information search screen of the patent document analysis software by performing a search with the specific patent document set as a search target and a search keyword with the cumulative patent classification in the column (3). The ratio of the number of documents by patent classification in the column (5) is a value obtained by dividing the number of documents in the column (4) by the total number of documents in the specific patent document set.

(Note 2) The ratio of the number of documents by cumulative patent classification of a specific relevant literature set refers to the patent classification used in the relevant literature increase search formula or the comprehensive relevant literature search formula from the patent classification used in the specific relevant literature set. It is an evaluation scale that acts as a sieve to be extracted, and is a column of the number of documents by cumulative patent classification in the table of the number of documents by cumulative patent classification of the specific relevant document set (for example, (5) in FIG. 4, FIG. 11, FIG. (See column).

SA3, created by layering keywords extracted from the master patent document set A for the relevant document increase retrieval formula A (Note 3) Using the hierarchical keyword table (Note 4), creating a hierarchical retrieval formula.

(Note 3) A keyword table classified by hierarchy of a specific relevant document set (for example, see FIG. 7, FIG. 12, FIG. 18) is a hierarchical expression that is a search expression constituting a patent document increase search expression or an exhaustive relevant document search expression. A table in which keywords extracted from the specific relevant literature set used in the search formula are stratified by hierarchy, and is composed of a plurality of hierarchical columns and a plurality of rows, and a specific patent literature item of the specific relevant literature set The keywords extracted for use in the patent document augmentation search formula and the exhaustive relevant literature search formula from among the keywords used in the above are all keywords expressing the category of the invention. All the keywords having the function of directly modifying one of the keywords in the first hierarchy column are arranged in each row, and each keyword in the second hierarchy column is assigned to any row in the second hierarchy column. The keywords extracted as keywords to be used in the relevant reference increasing search expression or the exhaustive relevant reference search expression, such as placing all of the keywords having the function of close modification in arbitrary rows of the third hierarchy column, respectively, For example, in the case of four keywords such as watch, device, elderly living alone, and life, the keyword expressing the category of the invention is based on the device and the device. Since the word order as Japanese is the device for the elderly living alone, the device is in the first hierarchy, the watch is in the second hierarchy, the life is in the third hierarchy, and the elderly Are arranged in the fourth hierarchical column, and the number of hierarchical columns can be arranged in any hierarchical column of keywords extracted from the specific relevant literature set for the relevant literature increasing search formula or the comprehensive relevant literature search formula. Yo If there are keywords (compound words) that belong to multiple hierarchical columns among the keywords to be placed and arranged in the keyword, the keywords are separated so that they can be arranged in a single hierarchical column, and the separated keywords are assigned to the corresponding hierarchy. If there is a partially matching keyword in the same hierarchy column, only the partially matching keyword is placed, and if there is a duplicate keyword, the keyword is placed without duplication.

A specific method for extracting keywords used in a specific patent document item of the specific relevant document set is to open the patent document analysis software and draw a ranking map creation screen on the display device 11, and on the drawn screen. The analysis target document set is the specific relevant document set, the data to be displayed is the number of documents by keyword, and the patent document item of the display target keyword is the name of the invention or the required patent document item such as an abstract with the mouse 8 When the selection / execution key is operated, a keyword ranking ranking map by keyword for the selected patent document item is drawn. Therefore, when the map is switched to a list by operating the mouse 8, the number of documents by keyword of the selected patent document item is displayed. Since the ranking table is obtained, the keyword ranking table created for each required patent document item is overlaid, and the overlapping table is created. The relevant document increase search formula from the keyword-specific document number table (see FIG. 5 as an example) for a specific patent document item in the specific relevant document set created by arranging the keywords in the keyword-specific document number table in ascending or descending order. And keywords to be used in an exhaustive relevant literature search formula.

When the number of keywords of a specific patent document item is large, the keywords determined to be used in the relevant document increase search formula or the exhaustive compatible document search formula are used as the number of documents by keyword of the specific patent document item (for example, A keyword is arranged in a hierarchical organization table (see FIG. 6 as an example) having the same number of rows and the same hierarchical column as the hierarchical keyword table of the specific relevant document set. If there is a matching keyword, it can be easily created by organizing it into a common part keyword, and if there is a duplicate keyword, organizing it into one keyword and creating a keyword table classified by hierarchy of the specific relevant document set.

(Note 4) Hierarchical search formula of a specific relevant document set is a type of search formula that constitutes a logical expression of a relevant reference increase search formula or an exhaustive relevant literature search formula, and is a keyword by hierarchy of the specific relevant literature set. For each hierarchical column of the table, the search item is a specific patent document item, the search keyword is a logical sum of keywords arranged in the same hierarchical column of the specific keyword table of the specific relevant document set, and the search method is a partial match search (See, for example, the formula in FIG. 8, K1, K2, K3, K4, the formula in FIG. 13, K1, K2, K3, K4, the formula in FIG. 19, K1, K2, K3, K4) The search formulas are combined with logical operators to increase the relevant literature search formula (see the relevant literature increase formula in FIG. 8 as an example) or the exhaustive relevant literature search formula (eg, the exhaustive relevant literature search in FIGS. 13 and 19 as an example). Applies to formula).

Patent classification extracted from the master patent document set A determined to be used in the conforming document increase search formula A in step SA4 and step SA2 (for example, G08B25 / in which the value in the column (5) in FIG. 4 satisfies the criterion) And the search reference for each hierarchy created in step SA3 (see, for example, the formulas in FIG. 8, K1, K2, K3, and K4). Browse).

The patent classification determined to be used in the relevant document increase search expression A in the above step is a logical sum (for example, the value of the column (5) in FIG. 4 satisfies the judgment criterion (3) the cumulative patent category of the column, G08B25 / Ref.) Applicable to the above-mentioned relevant document increase search formula A.

The method of applying the hierarchical search expression created in step SA3 to the relevant reference increase search expression A is a document including the relevant literature that is hit by not including it in the logical product of all hierarchical search expressions created in step SA3. Create a logical product of all hierarchical search expressions other than the hierarchical search expression for every hierarchical search expression for which you want to increase the number of documents in the set, and search for all the created logical products as logical sums An expression is created, and the search expression is applied as a keyword of the relevant document increase expression.

The method of applying the hierarchical search expression to the relevant document increase search expression A will be described in detail with reference to the hierarchical search expressions K1, K2, K3, and K4 in FIG. 8. K1 * K2 * K3 * K4 For example, if it is determined that the keyword used for K4 is insufficient in the search expression, the search expression of K1 * K2 * K3 not including the search expression K4 determined to be insufficient is used as the K1 * K2 If the method used in place of * K3 * K4 is taken, the number of nonconforming documents in the hit document set increases, but the number of conforming documents can also be increased, so the hierarchy in which the method is judged to be insufficient. The number of documents in the document set including the matching documents that are hit by not including the result of concisely describing the method of increasing the number of matching documents applied to all of the different search expressions in the logical product of the search expressions by all hierarchies Create a logical product of all hierarchical search formulas other than the hierarchical search formula for every hierarchical search formula to be increased, create a search formula that logically ORs all the created logical products, It is described that a search formula is applied as a keyword of the relevant literature increasing formula.

Regarding the selection of hierarchical search formulas that are not included in the logical product of the hierarchical search formulas in the hierarchical search formula A and that want to increase the number of relevant documents to be hit, Number of hits in the literature (for example, the number of hits in FIGS. 8 and 19 corresponds to the search of patent database in the case where the lower number of hits is not included in the logical product of the search expressions by hierarchy. The number of hits will increase).

SA5, a patent database search is performed using the relevant document increase search formula A, and a hit document set is a hit document set Ai (the initial value of the suffix i is 1 and the increment value is 1).

Specifically, the patent database 13 is connected to the patent analysis apparatus 1 via the communication line 14, and the literature search of the patent database 13 is performed by the relevant document increase search expression A by the operation of the input device 9. Is stored in the storage device 5, and the stored document set is converted into a data format that can be processed by the patent document analysis software and stored in the storage device 5.

SA6, using hit document set Ai which is the document set hit in the above step and master patent document set A, and if i is 1, it is a logical OR set of hit document set Ai and master patent document set A When the document increase set Ai is equal to or greater than 2, a document increase set Ai that is a logical sum set of the hit document set Ai and the document increase set A (i−1) is created.

The document search screen of the patent document analysis software is displayed on the display device 11 as a method of creating a document augmentation set Ai that is a logical sum set of the hit document set Ai and the master patent document set A or the document augmentation set A (i-1). If the creation of the logical sum set of the hit document set Ai and the master patent document set A or the document augmented set A (i-1) is specified and executed using the mouse 8 on the screen Can be created.

In step SA7, the document augmentation set Ai is screened, and a matching document augmentation set Ai is created.

The literature augmentation set Ai is screened by an arbitrary method, and the relevant literature augmentation set Ai is created from the extracted relevant literatures.

SA8, the comprehensiveness of the number of documents of the relevant document augmented set Ai is determined by sensory evaluation.

Perform a sensory evaluation, such as whether the patent document of interest is included in the relevant reference augmented set Ai, or whether the number of documents in the relevant reference augmented set Ai is suitable for the purpose of use. If the relevant reference augmented set Ai is the target comprehensive relevant patent literature set, and if it fails, the process returns to the start step and uses the relevant reference augmented set Ai instead of the master patent reference set A created in SA1. Then, each step is performed until the SA8 conformity document augmentation set Ai passes the sensory evaluation.

The relevant reference document augmented set Ai that has passed the sensory evaluation of SA9 and SA8 is the comprehensive comprehensive patent document set.

2. A method for searching patent documents related to claims 3 and 4 will be described below. This method is based on a scatter diagram for determining the number of documents satisfying (Note 5) of the master patent document set B (Note 6). This is an exhaustive search method for relevant documents when the number of cases can be obtained.

(Note 5) A scatter diagram for determining the number of documents in a specific relevant document set is used to determine whether or not the number of documents in the specific relevant document set is sufficient to extract keywords used in an exhaustive relevant document search expression. Is a scatter diagram of the number of documents and the number of keywords (for example, see FIGS. 10, 15, and 16), and the method of creating the scatter diagram arranges the documents of the specific relevant document set at random, The arranged documents are assigned i numbers (i = 1, 2, 3,..., N) continuously changing from 1 to N, and are composed of i documents from 1 to i. The number of keywords extracted for the purpose of use in the comprehensive reference retrieval formula extracted from each document set i in which i is changed is created and the keyword Draw a scatter diagram of the document i of the relevant document set i from which To create.

For example, referring to FIG. 10 (the number of documents in the master patent document set b is 277), 277 documents in the master patent document set b are randomly arranged using random numbers, and 1 to 277 are arranged in the arranged documents. The document set 20, the document set 50, the document set 100, and the document set 150 composed of the numbers 1 to 20, 1 to 50, 1 to 100, and 1 to 150 are created. The keywords used in the comprehensive matching document search formula are extracted from the table, the keyword table for each set is created, and the number of keywords in the keyword table is counted. The document set 20 is 31, the document set 50 is 50, the documents Since there are 56 sets 100 and 59 document sets 150, FIG. 10 is a scatter diagram drawn using these numerical values.

(Note 6) The number of documents that satisfy the keyword saturation point of a specific set of patent documents refers to the number of documents satisfying the number of documents in the specified set of relevant documents (see FIGS. 10, 15, and 16 as examples). Estimated that the number of keywords for which the increase in the number of keywords with respect to the increase in the number of documents corresponds to zero, in other words, the keywords for the comprehensive matching document search formula hardly increase even if the number of documents exceeding the keyword saturation point matching document increases. The number of documents to be obtained is determined by referring to the measurement points of the number of documents and the number of keywords on the scatter diagram for determining the number of documents in the specific relevant document set. You may obtain | require as the number of applicable literatures in which the gradient of a tangent becomes zero, and approximate the said scatter diagram with a quadratic curve (refer the quadratic curve in the figure of FIG. 10, FIG. 16 as an example), Fine Values may be determined as compatible document number becomes zero.

SB1, master patent document set B is created.

The creation method of the master patent document set B in this step may be any method. For example, a keyword extracted from the relevant patent document (see FIG. 9 as an example of a search expression using the keyword), or a patent classification investigated on the patent information platform There is a method in which a patent document search is performed using a search expression created using a keyword that has been thought of, and a hit literature set is screened and created by an arbitrary method.

A scatter diagram for determining the document number satisfaction degree of SB2 and master patent document set B (see FIG. 10 as an example) is created.

A scatter diagram for determining the document number satisfaction degree of the master patent document set B is created by the method of (Note 5).

SB3, the number of keyword saturation point matching documents in the master patent document set B is obtained.

The number of keyword saturation point matching documents in the master patent document set B is obtained by the method of (Note 6).

In the case of this item related to claim 2, since the number of documents in the master patent document set B is equal to or greater than the number of keyword saturation point matching documents, a direct comprehensive matching document search formula is created without going through the process of increasing the number of matching documents. Go to the step to do.

In SB4, the patent classification used in the comprehensive relevant document search formula B is determined by using the cumulative patent classification document number ratio in the cumulative patent classification document number table of the master patent document set B as an evaluation scale.

Using the same method as SA2, the patent classification to be used for the comprehensive relevant document search formula B is determined from the master patent document set B of SB1.

SB5, comprehensive matching document search from the document set created by randomly extracting from the master patent document set B the number of matching documents satisfying the reference number standard for the number of keyword documents in the master patent document set B (Note 7) Using a hierarchical keyword table (see FIG. 12 as an example) created by stratifying the keywords extracted for Formula B, a hierarchical search formula (examples such as the formulas in FIG. 13, K1, K2, K3, and K4) is created. .

(Note 7) The reference number criterion for the number of keywords in a specific relevant patent document set is a comprehensive value that is randomly extracted from the specific relevant document set for the number of keyword saturation point relevant documents in the specific relevant patent document set. A criterion value that uses the ratio of the number of documents in the document set as a source for extracting relevant document search formula keywords as an evaluation measure. Extracts comprehensive keywords for search criteria from a document set that satisfies the criteria value. If this is the case, it is considered that sufficient keywords can be extracted to create a search expression that can comprehensively collect relevant documents.

Theoretically, in order to extract 100% of the keywords used in the exhaustive relevant literature search formula, the keyword reference number satisfaction degree reference value of the literature set that is the source of extraction of the exhaustive relevant literature search formula keyword is 100% or more. However, since the slope of the approximate curve of the scatter diagram near the keyword saturation point of the relevant document is close to zero in the scatter diagram for determining the number of documents in the specific relevant patent document set, this is not practical. In consideration of this, the keyword document number sufficiency reference value may be determined according to the purpose of collecting relevant documents.

SB6, an exhaustive relevant document search formula B is created.

The patent classification determined to be used in the exhaustive relevant document search formula B in SB4 is a logical sum (for example, the cumulative patent classification in the column (3) where the value in the column (5) in FIG. 11 satisfies the criterion), G08B25 / reference) This is applied to the comprehensive relevant document search formula B.

The method of applying the hierarchical search expression created in step SB5 to the comprehensive relevant reference search expression B includes the relevant literature that is hit by not including it in the logical product of all hierarchical search expressions created in step SB5. Create a logical product of all hierarchical search expressions other than the hierarchical search expression for each hierarchical search expression for which the number of documents in the document set is to be increased, and make all the generated logical products be logical sums. A search formula is created, and the search formula is applied as a keyword of the relevant literature increasing formula.

SB7, a patent database search is performed using the comprehensive relevant document search formula B.

A patent database search is performed using the exhaustive relevant document search formula B in the same method as SA5.

A logical sum set of the document set hit in the patent database search of SB8 and SB7 and the master patent document set B is created, and the logical sum set is screened.

The document set of the relevant documents extracted by the screening of the logical sum set of SB9 and SB8 becomes the comprehensive relevant patent document set B.

3. The patent document search method related to claim 5 will be described below. This method is applicable when the number of keyword saturation point matching documents cannot be obtained from the scatter diagram for determining the number of documents in the master patent document set C. This is an exhaustive search method.

SC1, master patent document set C (see FIG. 3 as an example) is created.

The creation method of the master patent document set C in this step may be any method. For example, a search expression created using keywords extracted from relevant patent documents, or keywords that have come up with patent classifications investigated by the patent information platform. There is a method in which a patent document search is performed and a hit document set is screened and created by an arbitrary method.

SC2 creates a scatter diagram for determining the document number satisfaction degree of the master patent document set C.

A scatter diagram for determining the document number satisfaction degree of the master patent document set C (see FIG. 15 as an example) is created by the method of (Note 5).

SC3, the number of keyword saturation point matching documents in the master patent document set C is obtained.

Even if an attempt is made to determine the number of keyword saturation point matching documents in the master patent document set by the method of (Note 6), in the case of this paragraph relating to claim 5, the number of keyword saturation point matching documents in the master patent document set C is determined. In this case (for example, the master patent document set keyword saturation point matching document count cannot be obtained from FIG. 15), the comprehensive matching document search formula C is created through the process of increasing the matching documents. Go to step.

SC4, determine the patent classification to be used in the matching document increase search expression C using the ratio of the number of documents classified by cumulative patent classification in the cumulative patent classification document count table of the master patent document set C as an evaluation scale.

The number of documents by cumulative patent classification of the master patent document set C (see FIG. 4 as an example) is created by the method of (Note 1), and the ratio of the number of documents by cumulative patent classification in the document count table by cumulative patent classification (example) As shown in FIG. 4, column (5)) is used as an evaluation scale to determine the patent classification to be used in the relevant document increase search formula C.

SC5, the keywords used for the relevant document increase search formula C are extracted from the master patent document set C to create a hierarchical keyword table, and the hierarchical search formula is created using the hierarchical keyword table.

Create a keyword table classified by hierarchy (see FIG. 7 as an example) of the master patent document set C for determining keywords to be used in the relevant document increase search formula C by the method described in Note 3, and use the keyword table classified by hierarchy Thus, the hierarchical search formulas (examples K1, K2, K3, K4 in FIG. 8) are created by the method of Note 4.

SC6, Relevant literature increase search formula C is created.

The patent classification determined to be used in the relevant document increase search formula C in SC4 is a logical sum (for example, the value of column (5) in FIG. 4 satisfies the judgment criterion (3) cumulative patent classification in column (3), G08B25). / Reference) This is applied to the relevant document increase search formula C.

The method of applying the hierarchical search expression created in SC5 to the relevant reference increase search expression C is a document including the relevant literature that is hit by not being included in the logical product of all hierarchical search expressions created in step SC5. Create a logical product of all hierarchical search expressions other than the hierarchical search expression for every hierarchical search expression for which you want to increase the number of documents in the set, and search for all the created logical products as logical sums An expression is created, and the search expression is applied as a keyword of the relevant document increase expression.

SC7, a patent database search is performed using the relevant document increase search formula C, and the hit document set is set as a hit document set Ci (suffix, initial value of i is 1 and increment value is 1).

The patent database search is performed by the same reference document increasing search formula C by the same method as SA5.

SC8, using the document set Ci hit in the step and the master patent document set C, if i is 1, the document augmented set Ci that is the logical sum of the hit document set Ci and the master patent document set C is , I is equal to 2 or more, a document increase set Ci that is a logical sum set of the hit document set Ci and the document increase set C (i−1) is created.

The document search screen of the patent document analysis software is displayed on the display device 11 as a method of creating a document increase set Ci that is a logical sum of the hit document set Ci and the master patent document set C or the document increase set C (i-1). If the creation of a logical sum set of the hit document set Ci and the master patent document set C or the document augmented set C (i-1) is specified and executed using the mouse 8 on the screen Can be created.

SC9 screens the document augmentation set Ci to create a matching document augmentation set Ci.

The document augmentation set Ci is screened by an arbitrary method to create a matching document augmentation set Ci.

SC10, create a scatter diagram for determining the document number satisfaction degree of the relevant document augmentation set Ci.

A scatter diagram for determining the number of documents satisfying degree of the relevant document augmented set Ci is created by the method described in (Note 5) (see FIG. 16 as an example).
SC11, find the number of keyword saturation point matching documents in the matching document augmented set Ci.

The number of keyword saturation point matching documents in the matching document augmented set Ci is obtained by the method described in (Note 6).

If the number of documents in the relevant reference augmented set Ci is insufficient and the number of keyword saturation point relevant references cannot be obtained, the procedure returns to the start step and instead of the master patent literature set C created in SC1, the relevant reference augmented set The steps up to SC11 where the number of keyword saturation point matching documents in the matching document augmented set Ci is obtained using Ci are executed.

SC12 determines a patent classification to be used for the comprehensive relevant literature search formula C from the patent classification used in the relevant literature augmentation set Ci.

The document number table by cumulative patent classification (see FIG. 17 as an example) of the reference document augmented set Ci is created by the method of (Note 1), and the ratio of the number of documents by cumulative patent classification in the document number table by cumulative patent classification (example) As shown in FIG. 17, column (5)) is used as an evaluation scale to determine the patent classification to be used in the comprehensive relevant document search formula C.

SC13, a hierarchical keyword table created by stratifying keywords extracted for a comprehensive matching document search formula C from a document set having a document number satisfying the criterion for the number of keyword documents extracted at random from the matching document augmentation set Ci Create a hierarchical search expression using.

A keyword is extracted from a document set consisting of the number of documents satisfying the criterion for satisfying the number of keywords for keywords extracted at random from the matching document augmented set Ci, and a keyword is extracted from the document set using the comprehensive matching document search formula C. A separate keyword table (see FIG. 18 as an example) is created, and a search query by hierarchy of the relevant document augmented set Ci (see the formulas in FIG. 19, K1, K2, K3, and K4 as examples) is created from the hierarchical keyword table. To do.

SC14, to create the comprehensive relevant literature search formula C.

The patent classification determined to be used in the comprehensive relevant literature search formula C in SC12 is a logical sum (for example, the cumulative patent classification in column (3) where the value in column (5) in FIG. 17 satisfies the criteria) , G08B25 / reference) and applied to the comprehensive relevant document search formula C.

The method of applying the hierarchical search expression created in SC13 to the comprehensive relevant reference search expression C is a document including a matching reference that is hit by not including it in the logical product of the search expressions classified by hierarchy created in step SC13. Create a logical product of all hierarchical search expressions other than the hierarchical search expression for every hierarchical search expression for which you want to increase the number of documents in the set, and search for all the created logical products as logical sums An expression is created, and the search expression is applied as a keyword of the comprehensive matching document search expression C.

SC15, a patent database search is performed using the comprehensive relevant literature search formula C.

A patent database search is performed using the exhaustive relevant document search formula C in the same manner as SA5.

A conforming document exhaustive collection set C, which is a logical sum set of the document set hit in the patent database search of SC16 and SC15, and the relevant reference augmented set Ci is created, and the set is screened.

A document set of relevant documents extracted by screening the relevant document comprehensive collection set C of SC17 and SC16 becomes the comprehensive relevant patent document set C.

The following is an example of a patent document search method that increases the number of relevant documents using a collection of documents related to life monitoring such as safety confirmation, emergency notification, bathroom monitoring, etc. Three cases are described.

Example 1 relates to claims 1 and 2 in the case of performing a reference document search until the reference document augmentation set increased by using the patent classification and keywords of the master patent document set a passes the sensory evaluation. If the number of keyword saturation point matching documents in the master patent document set b regarding claims 3 and 4 can be determined, Example 3 determines the number of keyword saturation point matching documents in the master patent document set c regarding claim 5 It is a case where it is impossible.

Example 1 (Document retrieval method corresponding to claims 1 and 2 until the increased matched literature augmentation set passes sensory evaluation)

A1 is a step of creating a master patent document set a.

The master patent document set a was set to five relevant documents shown in FIG. 3 arbitrarily selected from the relevant documents searched using the patent information platform.

A2 is a step of determining a patent classification to be used in the matching document increase search formula a using the ratio of the number of documents classified by cumulative patent classification in the cumulative patent classification document count table of the master patent document set a as an evaluation scale.

FIG. 4 is a table of the number of documents by cumulative patent classification of the master patent document set a of this embodiment, and the ratio in the (5) cumulative patent classification by number of documents in the figure exceeds 80% of the criterion value. Since it is G08B25 /, the patent classification is determined as the patent classification to be used for the relevant document increase search formula a.

A3, a step of creating a hierarchical search formula using a hierarchical keyword table created by stratifying keywords extracted from the master patent literature set a for the relevant literature increase search formula a.

The number of documents in the master patent document set a in this embodiment is 5, and the total number of keywords used in the names, abstracts, and claims of the 5 inventions is 384. FIG. 7 is a hierarchical keyword table created by stratifying the keywords of the patent document items because they are within the capability range.

When the hierarchical keyword table (FIG. 7) is created, if the number of extraction target keywords is increased to 384, the name and summary of the invention of the master patent document set a as the extraction base and the keyword-specific documents in the claims Since keyword extraction is performed while looking at one part of the number table (Fig. 5), it is difficult to categorize keywords that are duplicated and compound words that need to be decomposed directly into the keyword table by hierarchy. Once the keywords are grouped into a hierarchical organization table (FIG. 6), if there are keywords that partially match in the same hierarchy column, they are integrated into the partially matching keywords and there are overlapping keywords. In this case, a keyword table for each hierarchy of the master patent document set a in FIG. 7 is created without duplication, and a search expression for each hierarchy (expressions in FIG. 8, K1, K2,. 3, K4) was created.

A4, using the patent classification extracted from the master patent document set a determined to be used in the relevant document increase search formula a in step A2 and the hierarchical search formula created in step A3, create the relevant document increase search formula a It is a step to do.

Since the patent classification determined for the relevant reference increase search expression a in step A2 is only G08B25 /, the patent classification is applied to the relevant reference increase search expression a in FIG.

The application method of the hierarchical search expression created in step A3 to the relevant reference increase search expression a (relevant reference increase search expression a in FIG. 8) is the logical product of all hierarchical search expressions, K1 * K2 * K3 * K4. The hierarchical search formulas that want to increase the number of documents in the document set including the matching documents that are hit by being not included in the list are the formulas in FIG. 8, K1, K2, K3, and K4, so that the hierarchy for each K1, K2, K3, and K4 Create logical products K1 * K2 * K3, K2 * K3 * K4, K3 * K4 * K1, and K4 * K1 * K2 between hierarchical search formulas that do not include another search formula, and the logical sum of the logical products, (K1 * K2 * K3 + K2 * K3 * K4 + K3 * K4 * K1 + K4 * K1 * K2) was applied to the relevant literature increase search formula a.

A5, which is a step of performing a patent database search by the conforming document increasing search formula a, and setting a hit document set as a hit document set ai (suffix, i having an initial value of 1 and an increment value of 1).

This step is for i = 1.

As a result of searching for patent documents using the communication line 14 and connecting to the patent database 13 and performing the patent document search using the relevant document increasing search formula a, i in this step is 1, so hit documents The set ai became the hit document set a1, and the number of documents in the document set was 1173.

Since A6 and i are 1, this is a step of creating a document augmentation set a1 which is a logical sum set of the hit document set a1 which is the document set hit in A5 and the master patent document set a.

The number of documents in the hit document set a1 is 1173, and since all the five documents in the master document set a in this embodiment are included in the hit document set a1, the number of documents in the document increase set a1 is 1173. It was.

A7 is a step of performing screening of the literature augmentation set a1 to create a relevant literature augmentation set a1.

1173 patent documents in the document augmentation set a1 were screened and 607 relevant documents were extracted. Therefore, the number of literatures in the relevant literature augmentation set a1 was 607.

A8 is a step of sensory evaluation of the relevant document augmented set a1.

The number of documents in the relevant literature augmented set a1 is 607, the master patent literature set a is five, and all the five patent documents are included in the relevant literature augmented set a1, and the number of the five literature references. Since the number of documents in the relevant literature augmented set a1 is 121 times, the relevant literature augmented set a1 is considered acceptable.

A9, The relevant literature augmented set a1 becomes the target comprehensive relevant patent literature set.

Example 2 (corresponding to claims 3 and 4, when the number of keyword saturation point matching documents in the master document set b can be obtained)

B1 is a step of creating a master patent document set b.

The master patent document set b of Example 2 does not use patent classification, but creates a search formula using only the keywords that come to mind regarding watching, FIG. 9 is performed, a patent database search is performed using the search formula, and 490 hits are screened. 277 relevant documents were extracted, and the extracted document set was designated as a master patent document set b.

B2 is a step of creating a scatter diagram for determining the document number satisfaction degree of the master document set b.

FIG. 10 is a scatter diagram for determining the document number satisfaction degree of the master document set b.

B3, a step of obtaining the number of keyword saturation point matching documents in the master patent document set b.

The quadratic equation in FIG. 10 is an approximate curve of the scatter diagram in the figure, and the number of relevant documents whose derivative value of the number of keywords with respect to the number of relevant documents is zero from the approximate curve (the number of relevant documents is 0.0026 * 2 * X = 0.6349, X) is 122, so the number of keyword saturation point matching documents in the master patent document set b can be determined to be 122, so the process proceeds to step B4.

B4, a step of determining a patent classification to be used in the comprehensive matching document search formula b using the ratio of the number of documents classified by cumulative patent classification in the cumulative patent classification document count table of the master patent document set b as an evaluation measure.

The document number table by cumulative patent classification of the master patent document set b is as shown in FIG. 11, and in the figure, the patent classification satisfying the judgment criteria in the (5) document number ratio column by cumulative patent classification is NO1. Since it is G08B25 /, the patent classification is used as the patent classification used for the comprehensive relevant document search formula b.

B5, hierarchies created by stratifying keywords extracted for a comprehensive matching document search expression b from a document set having a document number satisfying the reference number satisfaction degree standard value for keywords extracted randomly from the master patent document set b This is a step of creating a hierarchical search expression using a keyword table.

If the keyword document number satisfaction level reference value of the second embodiment is 100% or more, the number of keyword saturation point matching documents in the master patent document set b is 122 from B3, so the keywords by hierarchy of the comprehensive matching document search formula b are as follows. Since the number of documents in the document set for extracting the keywords in the table needs to be 122 or more, in Example 2, the keywords of the document set created by randomly extracting 150 documents from the master patent document set b are used. 12 was used to create a hierarchical keyword table in FIG. 12, and a hierarchical search formula (formulas in FIG. 13, K1, K2, K3, K4) was created using the hierarchical keyword table.

B6 is a step of creating an exhaustive relevant document search formula b.

Since the patent classification decided to be used for the comprehensive relevant document search expression b in B4 is G08B25 /, it is created by using the patent classification and the keyword of the hierarchical keyword table created for the comprehensive relevant reference expression b in B5 The comprehensive reference retrieval formula b was created by combining the hierarchical retrieval formulas K1, K2, K3, and K4 in FIG.

The combination of hierarchical search formulas K1, K2, K3, and K4 applied to the comprehensive relevant document search formula b in FIG. 13 is a search formula by all hierarchies for all of the hierarchical search formulas K1, K2, K3, and K4. In order to increase the number of documents in a document set including relevant documents by not including them in the logical product of each other, the hierarchical search expressions that do not include the hierarchical search expressions for each hierarchical search expression K1, K2, K3, K4 And (K1 * K2 * K3, K2 * K3 * K4, K3 * K4 * K1, K4 * K1 * K2) are created, and the logical sum of the logical products is represented in the comprehensive relevant literature search formula b in FIG. Applied.

B7 is a step of performing a patent database search using the comprehensive relevant document search formula b.

As a result of connecting to the patent database 13 using the communication line 14 and performing a patent document search using the comprehensive relevant document search formula b, 8401 documents were hit.

B8 is a step of creating a conforming document exhaustive collection set b that is a logical sum of the hit literature set and the master patent document set b, and screening the conforming literature exhaustive collection set b.

The number of documents in the relevant document comprehensive collection set b in Example 2, which is a logical sum set of 8401 hit document sets and 277 master patent document sets b, is 8456, and 8401 of the hit document sets. 55 cases are increased compared to

If all of the 277 relevant documents in the master patent document set b are included in the hit document set, the number of documents in the relevant document comprehensive collection set b should not increase, but it has increased by 55. The reason is that the FI of formula number (1) in the search formula of FIG. 13 is applied to G08B25 / in which the ratio of the number of documents by cumulative patent classification is 80% in FIG. 11 to the comprehensive reference search formula b of FIG. Therefore, the document set that is hit by the search formula does not include 55 matched documents corresponding to 20% of the 277 matched documents. In order to increase the ratio of the documents in the master patent document set included in the document set that hits in step 1, the criterion for determining the ratio of the number of documents by cumulative patent classification in the document count by cumulative patent classification of the master patent document set is 1 It can be seen that may be brought closer to 0%.

In order to estimate the number of relevant documents included in 8456 of the relevant literature comprehensive collection set b with a sampling method with an accuracy of ± 10%, 100 sample literatures are randomly selected from the relevant literature exhaustive collection set b. (See FIG. 14), the ratio of the relevant documents included in the extracted document set was 17%, so is the number of relevant documents included in the relevant document comprehensive collection set b equal to 1294? The number of cases is estimated to be equal to or less than or equal to 1581.

B9, the target comprehensive literature reference set b is the relevant literature set equal to or equal to 1294 extracted by the screening, or equal to or less than 1581.

Example 3 (corresponding to claim 5, when it is not possible to determine the number of keyword saturation point matching documents in the master document set c)

C1. This is a step of creating a master patent document set c.

The master patent document set c of Example 3 was set to five matching document sets shown in FIG. 3 which are the same as the document set used in Example 1.

C2 is a step of creating a scatter diagram for determining the number of documents in the master patent document set c.

FIG. 15 is a scatter diagram for determining the document number satisfaction degree of the master document set c.

C3, a step of obtaining the number of keyword saturation point matching documents in the master document set c.

Since the number of saturation point conformity documents cannot be obtained from the scatter diagram for determining the document number satisfaction degree determination of the master document set c of Example 3 and FIG. 15, in order to increase the number of conformity documents of the master document set c, the following Perform the steps.

C4 is a step of determining a patent classification to be used in the matching document increase search formula c using the ratio of the number of documents classified by cumulative patent classification in the cumulative patent classification document count table of the master patent document set c as an evaluation scale.

Since the master patent document c in this step is the same as that in the first embodiment, the patent classification used in the relevant document increase search expression c is G08B25 / which is the same as the patent classification extracted from FIG.

C5, a step of creating a hierarchical search expression using a hierarchical keyword table created by stratifying keywords extracted from the master patent document set c for the relevant document increase search expression c.

Since the master patent document of this step is the same as that of the first embodiment, the hierarchical keyword table of this embodiment is the above-described FIG. 7 and the hierarchical search expressions are the expressions in FIG. 8, K1, K2, K3, and K4. .

C6 is a step of creating a matching document increase search formula c.

Since the master patent document c in this step is the same as that in the first embodiment, the corresponding document increase search formula c in this embodiment is the same as the A4 reference document increase search formula c in FIG.

C7, a patent database search is performed using the relevant document increase search formula c, and a hit document set is a hit document set i (i is a suffix, i is an initial value of 1, and an increment value is 1).

Since the matching document increase search formula c of the present embodiment is the same as that of the first embodiment, the hit document set becomes 1173 document sets which are the same as the hit document set 1 of A5.

C8, this step is a step of creating a document augmentation set c1 which is a logical sum set of the hit document set c1 which is the document set hit in C7 and the master patent document set c because the suffix i is 1.

The number of documents in the document augmentation set c1, which is a logical sum of the hit document set c1 (the number of documents 1173) and the master document set c (the number of documents 5), is 1173.

C9 is a step of performing screening of the literature augmentation set c1 to create the relevant literature augmentation set c1 of Example 3.

1173 documents in the document augmented set c1 were screened and 607 matched documents were extracted. Therefore, the number of documents in the matched document augmented set c1 was 607.

C10 is a step of creating a scatter diagram for determining the number of documents satisfying the relevant document increase set c1.

FIG. 16 shows a scatter diagram for determining the degree of satisfaction of the number of keyword documents created using 607 of the relevant document augmented set c1 created in C9.

C11, a step of obtaining the number of keyword saturation point matching documents in the matching document augmentation set c1.

The keyword saturation point matching document number of the reference document augmented set c1 is an approximate curve of the scatter diagram of FIG. 16, and the number of documents whose differential value of the keyword number is zero with respect to the document number of the secondary approximate curve in the figure (the number of documents) Is obtained as 116 of 0.0018 * 2 * X = 0.4184). Therefore, the relevant reference augmented set c1 having 607 relevant literatures is the keyword of the comprehensive relevant literature search formula c. Since it is confirmed that the number of documents is sufficient to extract, the process proceeds to the following steps.

C12, a step of determining, as an evaluation measure, a patent classification used in the comprehensive relevant literature search formula c as a rating scale based on the cumulative patent category literature count ratio in the cumulative patent category literature count table of the relevant literature augmented set c1.

FIG. 17 is a table of the number of documents classified by cumulative patent classification of the relevant document augmented set c1 of Example 3, and (5) the ratio of the number of documents classified by cumulative patent classification satisfies 80% of the criterion value in the figure. Since the patent classification is G08B25 /, the patent classification is determined as the patent classification to be used for the exhaustive relevant document search formula c.

C13, a hierarchical keyword created by stratifying keywords extracted for a comprehensive relevant document search expression from a document set having a document number satisfying the reference number satisfaction degree standard value for keywords extracted at random from the relevant reference augmentation set c1 This is a step of creating a hierarchical search expression using a table.

If the keyword document number satisfaction degree reference value of Example 3 is set to 100% or more, the number of keyword saturation point matching documents in the matching document augmented set c1 is 116. Since the number of documents in the document set for extracting the keywords in the table needs to be 116 or more, in Example 3, the keywords of the document set created by randomly extracting 150 documents from the relevant document augmented set c1 are used. 18 was used to create a hierarchical keyword table, and a hierarchical search formula (formulas in FIG. 19, K1, K2, K3, K4) was created using the hierarchical keyword table.

C14, a step of creating the exhaustive relevant document search formula c.

Since the patent classification determined to be used for the comprehensive relevant document search formula c in C12 is G08B25 /, the patent classification and the keywords of the hierarchical keyword table created for the comprehensive relevant literature search formula c in C13 are used. The relevant document exhaustive search formula c was created by combining the hierarchical search formulas K1, K2, K3, and K4 in FIG.

The combination of hierarchical search formulas K1, K2, K3, and K4 applied to the comprehensive relevant literature search formula c in FIG. 19 is not included in the logical product of all hierarchical search formulas, K1 * K2 * K3 * K4. The hierarchical search formulas that want to increase the number of documents in the document set including the matching documents that hit are the same as A4 in the formulas in FIG. 19, K1, K2, K3, and K4. This was applied to the comprehensive relevant literature search formula c (FIG. 19).

C15 is a step of performing a patent database search using an exhaustive relevant document search formula c.

As a result of connecting to the patent database 13 using the communication line 14 and performing a patent document search using the comprehensive relevant document search formula c, 8154 documents were hit.

C16, a step of creating a relevant document comprehensive collection set c that is a logical sum of the hit literature set c and the relevant literature augmented set c1, and screening the relevant literature comprehensive collection set c.

The number of documents in the comprehensive document collection set c of the third embodiment, which is a logical sum of 8154 in the hit document set c and 607 in the matched document augmented set c1, is 8185, and the hit document set This is an increase of 31 cases compared to 8154 cases of c.

If all the 277 relevant documents in the relevant literature augmented set 1c are included in the hit literature set c, the number of documents in the relevant literature comprehensive collection set c in Example 3 should not increase, but 31. The reason for the increase is that the patent classification of the relevant reference augmented set c1 of the extraction base of the patent classification of the comprehensive relevant reference search formula c is the number of documents by cumulative patent classification in the master patent reference set c in FIG. 5) G08B25 / with 100% cumulative number of documents classified by patent category is used, but the G08B25 / (5) cumulative number of documents classified by patent category is 100%. The number of documents in the patent document set is as few as 5, and it is considered that the patent documents other than G08B25 / are not included in the patent document written together with G08B25 /. If the number of documents in the G08B25 / is much larger than 5, the ratio of the number of documents by G08B25 / will be less than 100%, and the corresponding documents to which patent classifications other than the G08B25 / are assigned will be hit. It is estimated that

In order to estimate the number of relevant literatures included in 8185 cases of the relevant literature comprehensive collection set c with an accuracy of ± 10% by the sampling method, 100 sample literatures are randomly selected from the relevant literature exhaustive collection set c. (See FIG. 19), the ratio of the relevant documents included in the extracted document set was 18%, so is the number of relevant documents included in the relevant document comprehensive collection set c equal to 1326? The number of cases is estimated to be equal to or less than 1621 cases.

C17, The relevant literature set c that is equal to or equal to 1326 extracted by the screening, or equal to or less than 1621 is the comprehensive comprehensive patent literature set c.

According to the patent document search method of the present invention, even a person unfamiliar with the patent classification and keywords of the search target technology and synonyms of the keyword can collect the patent documents in the search target field comprehensively. Since more people can take charge of tasks that are limited to those with specialized knowledge, it is easy to level the load on the parties concerned.

The invention described in claim 1 is a patent document search method executed by a computer in which patent document analysis software and spreadsheet software are installed.
A document set creation step of creating a document set X by randomly extracting the number of documents satisfying the keyword document number satisfaction degree reference value from the master patent document set,
An exhaustive relevant literature search formula creation step for creating an exhaustive relevant literature search formula using the patent classification used in the master patent literature set and a keyword extracted from the literature set X, and the exhaustive relevant literature search formula patent documents and acquiring an exhaustive compatible document set by screening the logical union of the master Patent document set and document set hit in the patent database searching step and patent database search step of a patent database search using This is the search method.

In the invention according to claim 2, in the exhaustive relevant document search formula, a search query by hierarchy is created for every hierarchy of the keyword table by hierarchy of the patent document set X. Create a logical product of all the hierarchical search expressions other than the hierarchical search expression for every hierarchical search expression for every hierarchical search expression that wants to increase the number of documents in the document set including the hit documents that are not included in the logical product, Create a search formula that logically ORs all the created logical products, and determine the search formula as a keyword for the comprehensive relevant document search formula and the ratio of the number of documents by cumulative patent classification of the master patent document set as an evaluation scale The patent document search method according to claim 1 , wherein the patent classification is used.

The invention according to claim 3 is created by randomly extracting from the master patent document set the number of documents that satisfy the keyword document number sufficiency standard value because the number of documents in the master patent document set is insufficient. When the document set X cannot be created , the keyword saturation point matching document number cannot be obtained by the scatter diagram for determining the document number satisfaction degree of the master patent document set by an arbitrary method . , the literature number of master document set is increased until the are determined keyword saturation point adapted Document number in any way, by using the adapted set of documents obtained by the increase as the master Patent document set, the document grouping production step, the patent database searching step, and executes a step of screening A patent document search method according to Motomeko 1.

Claims (5)

A patent document search method executed on a computer on which patent document analysis software and spreadsheet software are installed,
A matching document increasing search formula creating step for creating a matching document increasing search formula using the patent classification and keywords used in the master patent document set, and a patent database searching step for searching a patent database using the matching document increasing search formula; Screening a logical sum set of the document set hit in the patent database search step and the master patent document set to create an all-matched document set creating step and evaluating the all-matched document set by sensory evaluation If it passes, the above-mentioned all relevant document set is set as an exhaustive relevant document set,
If the all relevant document set is evaluated by sensory evaluation and rejected, the number of documents in the master patent document set is increased by an arbitrary method, and the relevant document increases until the all relevant document set passes by sensory evaluation. Performing a search formula creation step, the patent database search step, and the all relevant document set creation step;
A method for searching patent documents, wherein a search for relevant documents is continued until the set of all relevant documents is passed through sensory evaluation. In the relevant reference increasing search expression creation step, a search expression by hierarchy is created for every hierarchy in the keyword table by hierarchy of the master patent document set, and it is not included in the logical product of the created search expressions by all hierarchies. For each search expression classified by hierarchy for which the number of documents in the document set including the relevant documents to be increased is increased, a logical product of all the search expressions classified by hierarchy other than the search expression classified by hierarchy is created, and all of the created logical products Is created as a logical OR, and the search formula is a keyword of the relevant literature increasing formula, and the patent classification of the relevant literature increasing formula is determined by using the ratio of the number of documents by cumulative patent classification of the master patent literature set as an evaluation scale. The method for searching a patent document according to claim 1, wherein: A patent document search method executed on a computer on which patent document analysis software and spreadsheet software are installed,
A document set creation step of creating a document set X by randomly extracting the number of documents satisfying the keyword document number satisfaction degree reference value from the master patent document set,
An exhaustive relevant literature search formula creation step for creating an exhaustive relevant literature search formula using the patent classification used in the master patent literature set and a keyword extracted from the literature set X, and the exhaustive relevant literature search formula Patent document search using patent database search step, and performing comprehensive exhaustive document search by screening a logical sum set of a document set hit in the patent database search step and the master patent document set retrieval method. In the comprehensive relevant literature search formula, a hit-by-hierarchical search formula is created for every hierarchy of the hierarchical keyword table of the patent document set X and is not included in the logical product of the created hierarchical search formulas. Create a logical product of all hierarchical search expressions other than the hierarchical search expression for every hierarchical search expression for which the number of documents in the document set including the relevant documents is to be increased, and logically calculate all the generated logical products. Create a search formula that is summed, and use the search formula as a keyword of the exhaustive relevant literature search formula and a patent classification that is determined by using the cumulative number of patent classifications of the master patent literature set as an evaluation measure. The patent document search method according to claim 3, wherein the method is a search method. It is not possible to create the document set X in which the number of documents in the master patent document set is insufficient and the number of documents satisfying the keyword document number satisfaction standard is randomly extracted from the master patent document set. In this case, the number of documents in the master document set is increased by an arbitrary method until it exceeds the number of documents satisfying the keyword document number sufficiency standard value, and the increased matching document set is used as the master patent document set. 4. A method for searching patent documents according to claim 3, wherein exhaustively relevant documents are searched.

Images