JP2019057068A - Information processor and computer program - Google Patents

Abstract

To provide an information processor adapted to assist an alliance between organizations.SOLUTION: An information processor includes an alliance-destination search device 16 that extracts, on respective ones of a plurality of organizations, problems described in a technical document group having an association with an organization stored in a technical document DB 14. The alliance-destination search device 16 extracts, on respective ones of the plurality of organizations, elemental technologies for solving the problems described in the technical document group. The alliance-destination search device 16 classifies, for each of the organizations, the technical document group based on a combination of the problems and the elemental technologies. The alliance-destination search device 16 determines at least two organizations different in classification results as alliance candidates and outputs information about the alliance candidates to a user device 12.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a data processing technique, and more particularly to a technique for supporting partnership between organizations.

  In recent years, open innovations that increase market opportunities by utilizing inflows of resources such as technologies and ideas and expanding innovations created within the organization outside the organization have been actively performed. In addition, so-called “corporate matching” is being conducted in which organizations with different backgrounds seek out business partners outside the organization in order to create new businesses. In addition, a service (also referred to as “company matching service”) for searching for a partner of a company has been provided by a local government or a specialized private trader (see, for example, Patent Documents 1 and 2).

Japanese Patent Laying-Open No. 2015-228153 JP 2011-2322892 A

  The conventional company matching service manually searches for other companies that are candidates for cooperation with a certain company, and may take time.

  The present invention has been made in view of these problems, and one object is to provide a technology that supports partnerships between organizations.

  In order to solve the above-described problem, an information processing apparatus according to an aspect of the present invention includes a problem extraction unit that extracts a problem described in a technical document group related to an organization for each of a plurality of organizations, and a technical document group A technology extraction unit that extracts elemental technologies for solving problems described in the above for each of a plurality of organizations, and a classification unit that classifies technical document groups for each organization based on a combination of the issues and the elemental technologies. A determination unit that determines at least two organizations with different classification results by the classification unit as affiliation candidates, and an output unit that outputs information on the affiliation candidates determined by the determination unit.

  It should be noted that any combination of the above components and the expression of the present invention converted between a method, a program, a recording medium storing the program, and the like are also effective as an aspect of the present invention.

  According to the present invention, partnership between organizations can be supported.

FIGS. 1A to 1D are diagrams schematically showing a complementary relationship pattern between companies. It is a figure which shows the structure of the corporate alliance support system of 1st Example. It is a block diagram which shows the function structure of the cooperation partner search apparatus of FIG. It is a figure which shows typically the cross analysis result of one company. FIGS. 5A to 5D are diagrams illustrating an example of derivation of the score of pattern 1. 6A to 6D are diagrams illustrating an example of derivation of the score of pattern 1. FIG. FIGS. 7A to 7D are diagrams illustrating an example of derivation of the pattern 2 score. 8A to 8B are diagrams illustrating an example of derivation of the score of pattern 2. FIG. FIGS. 9A to 9D are diagrams illustrating an example of derivation of the score of pattern 3. FIGS. 10A to 10C are diagrams illustrating an example of derivation of the score of the pattern 4. FIGS. 11A to 11D are diagrams showing an example of derivation of the score of pattern 1 in the second modification. 12A to 12D are diagrams illustrating an example of derivation of the score of pattern 1 in the second modification.

  An outline will be described before describing the configuration of the business alliance support system of the embodiment. The business tie-up support system according to the embodiment automatically searches for and proposes a large company that can be a business partner to a small and medium-sized business looking for a business tie-up partner. Specifically, technical documents such as patents owned by each company are analyzed, and large companies that have a technical complementary relationship with SMEs are proposed as tie-up candidates.

  For small and medium-sized enterprises, it is said that a large enterprise that possesses technology complementary to the technology owned by the small and medium enterprise is likely to be a partner for joint research and development. The partner search apparatus of an Example performs the cross analysis centering on a subject and the elemental technology which solved the subject based on the patent document of each SME and large enterprise. As a result of cross-analysis, the partner search device considers that there is a complementary relationship when both of the issues and elemental technologies differ between the small and medium-sized enterprises and the large enterprises. Select.

  FIGS. 1A to 1D schematically show a complementary relationship pattern between companies. In FIGS. 1A to 1D, problems are arranged horizontally, and elemental technologies that solve each problem are vertically arranged. For example, FIG. 1A shows that a small and medium enterprise has solved the problem B by using the element technology X, and a large company has solved the problem A by using the element technology X. Pattern 1 shown in FIG. 1A is a pattern in which small and medium enterprises and large enterprises solve different problems with the same technical elements. When the SME and large company have a pattern 1 relationship, the partner search apparatus selects the large company as a candidate for the alliance because the merits of the alliance are large (possibly leading to a new business).

  Pattern 2 shown in FIG. 1B is a pattern in which a small and medium enterprise and a large enterprise solve the same problem with different elemental technologies. Moreover, the pattern 3 shown in FIG.1 (c) is a pattern in which the SME and the large company solved different problems with different technologies. The partner search apparatus selects a large company as a candidate for a partner because the merit of the partner is large even when the SME and the large company have a pattern 2 or 3 relationship.

  Pattern 4 shown in FIG. 1 (d) is a pattern in which a small and medium enterprise and a large enterprise solve the same problem with the same elemental technology. In this case, it is assumed that SMEs and large companies are in a competitive relationship, and that it is necessary to adjust rights. Therefore, the partner search device does not select a large company as a partner candidate because the merit of the partner is small for small and medium-sized companies.

  In addition, the partner search device of the embodiment can be applied to other than the matching between small and medium-sized enterprises and large enterprises, for example, various types of associations such as associations other than enterprises, foundations, public institutions, departments within enterprises, etc. Applicable for organization matching. In addition, the partner search apparatus of the embodiment analyzes the technology possessed by each company based on patent documents, but based on technical documents other than patent documents, such as academic papers, technical explanations, product manuals, white papers, etc. May be analyzed.

(First embodiment)
FIG. 2 shows the configuration of the corporate alliance support system 10 of the first embodiment. The corporate alliance support system 10 includes a user device 12, a technical document DB 14, and a partner search device 16. These devices are connected via a communication network 18 including a LAN, a WAN, and the Internet.

  The user device 12 is an information processing device of a company that considers business tie-ups with other companies (in the embodiment, a small and medium company, hereinafter also referred to as “analyzed company”), and is an employee of the analyzed company (hereinafter “user”). Is called). The user device 12 may be a PC, a tablet terminal, or a smartphone.

  The technical document DB 14 is a database server that stores technical documents of a plurality of companies that can be analyzed by the partner search device 16. The technical document in the embodiment is a patent document related to each company, for example, an open patent publication or a patent publication relating to a patent filed by each company. The partner search device 16 is an information processing device that automatically searches for companies that are candidates for the partner of the analyzed company.

  FIG. 3 is a block diagram showing a functional configuration of the partner search device 16 of FIG. The partner search device 16 includes a search request receiving unit 20, a document acquiring unit 22, a population specifying unit 24, a target company specifying unit 26, an issue extracting unit 28, a technology extracting unit 30, a document classifying unit 32, a difference determining unit 34, A tie-up candidate determination unit 36, a technology range specifying unit 38, a rank determination unit 40, and a tie-up candidate output unit 42 are provided. Although not shown, the partner search device 16 includes a communication unit that communicates with an external device according to a predetermined communication protocol. Each functional block in FIG. 3 transmits and receives data to and from an external device via a communication unit.

  Each block shown in the block diagram of the present specification can be realized in terms of hardware by an element or a mechanical device such as a CPU / memory of a computer, and can be realized by a computer program or the like in terms of software. Then, the functional block realized by those cooperation is drawn. Those skilled in the art will understand that these functional blocks can be realized in various forms by a combination of hardware and software.

  For example, the plurality of functional blocks in the partner search device 16 of FIG. 3 may be implemented as a computer program including a plurality of modules corresponding to these functional blocks. This computer program may be installed in the storage of the partner search device 16 via a recording medium or a network. The CPU of the partner search device 16 may exhibit the function of each functional block shown in FIG. 3 by reading this computer program into the main memory and executing it.

  The search request receiving unit 20 receives data (hereinafter also referred to as “search request”) that is transmitted from the user device 12 and requests search for partner candidates. The search request includes identification information of the analyzed company (for example, a small and medium company in which the user works). The document acquisition unit 22 acquires a group of patent documents related to the analyzed company from the technical document DB 14, and acquires, for example, one or more patent documents in which the analyzed company is set as an applicant or a patentee.

  The population specifying unit 24 extracts about 1000 patent documents similar to the patent documents of the analyzed company from the technical document DB 14 and specifies them as the analysis population. A similar natural language processing technique may be used for the similar determination. For example, the population specifying unit 24 uses the feature data obtained by vectorizing each of one or more patent documents acquired by the document acquiring unit 22 based on the described words, and other patent documents (in other words, the analyzed company). The inner product value may be calculated with feature data obtained by vectorizing a patent document related to the above) based on the described word. Then, the higher the inner product value, the higher the similarity may be determined. Note that feature data obtained by vectorizing each patent document may be stored in advance in the technical document DB 14.

  The target company specifying unit 26 specifies about five large companies as analysis targets in descending order of similarity from the analysis population specified by the population specifying unit 24. Hereinafter, large enterprises to be analyzed are also referred to as “target enterprises”. The scope of a large company may be determined in advance according to capital, the number of employees, and the like. For example, when the company that holds the patent document with the first similarity is not a large company, and the company that has the patent documents with the second to sixth similarities is a large company, the target company specifying unit 26 You may exclude the company with the 1st similarity and specify five large companies with the 2nd to 6th similarity as the target companies.

  The problem extraction unit 28, for one or more patent documents of the analyzed company (in the embodiment, a small and medium company) acquired by the document acquisition unit 22, issues to be solved by technology from the texts of each patent document. Extract. Further, the task extraction unit 28 acquires one or more patent documents of each of a plurality of target companies (large companies in the embodiment) from the population specified by the population specification unit 24. The problem extraction unit 28 extracts, for each target company, a problem to be solved by technology from the texts of each patent document related to the target company. The problem may be, for example, “size”, “weight”, “usability”, and the like.

  The technology extraction unit 30 extracts elemental technologies that have solved the problem from the texts of each patent document for one or more patent documents of the analyzed company acquired by the document acquisition unit 22. Further, the technology extraction unit 30 acquires one or more patent documents of each of the plurality of target companies from the population specified by the population specification unit 24. The technology extraction unit 30 extracts, for each target company, elemental technologies that have solved the problem from the text of each patent document. The elemental technology may be, for example, “pressure control”, “temperature control”, “density control”, or the like.

  The task extracting unit 28 and the technology extracting unit 30 may extract the task and the element technology described in the patent document by a known text mining technology. For example, the problem extraction unit 28 and the technology extraction unit 30 describe predetermined items (for example, “international patent classification”, “technical field”, “problem to be solved by the invention”, “means for solving the problem”) in patent documents. Issues and elemental technologies may be extracted based on the content. The issue extraction unit 28 and the technology extraction unit 30 are grouping analysis functions provided by NIR Cyber Patent Co., Ltd. (https://www.patent.ne.jp/service/macro/bunseki_grouping.html) May be used to extract problems and elemental technologies described in each patent document.

  The document classification unit 32 classifies one or more patent documents (patent document group) of each company based on the combination of the problem and the element technology for each company with respect to the analyzed company and the target company. The document classification unit 32 according to the embodiment performs a cross analysis that classifies each of a plurality of patent documents of each company on two axes of a task and an elemental technology. In other words, the document classification unit 32 determines which combination of a plurality of types of combinations of issues and elemental technologies each of a plurality of patent documents of each company corresponds to.

  The document classification unit 32 generates data (hereinafter also referred to as “cross analysis result”) in which the classification result of the patent document group of each company is expressed by a two-dimensional chart with the subject and the elemental technology as axes. FIG. 4 schematically shows a cross analysis result of one company. In the cross analysis result 50 of FIG. 4, the tasks are arranged horizontally and the elemental technologies are arranged vertically. In the cross analysis result 50, the number near the intersection (upper left) between one problem and one element technology indicates the number of patent documents in which the problem and the element technology are described.

  The number-of-documents symbol 52 is a symbol for making it easy to visually grasp the number of documents, and is provided at the intersection of an issue and elemental technology. In FIG. 4, the document number symbol 52 is added only to the intersection of accuracy / accuracy and pressure control. However, the document classification unit 32 performs the document number symbol 52 for each combination of the problem and the element technology (that is, each intersection). May be added. In addition, the document classification unit 32 increases the size of the document number symbol 52 added to the intersection indicating the combination as the number of documents in the combination of the problem and the element technology increases.

  Returning to FIG. 3, the difference determination unit 34 includes a combination of the analyzed company and the first target company, a combination of the analyzed company and the second target company,..., A combination of the analyzed company and the Nth target company. Each time, the classification result of the document classification unit 32 is determined to be different. In the first embodiment, if the classification result obtained by the document classification unit 32 has the same combination (hereinafter also referred to as “focused technical field”) having the largest number of patent documents belonging to a plurality of combinations of problems and elemental technologies. It is determined that the classification results are the same. On the other hand, if the focused technical field is different, it is determined that the classification result is different. In the example of FIG. 4, the focus technical field is a combination of the problem “accuracy / accuracy” and the elemental technology “pressure control”. The difference determination unit 34 further determines whether any of the patterns 1 to 3 in FIG.

  The affiliation candidate determination unit 36 determines at least two companies with different classification results by the document classification unit 32 as affiliation candidates. In the embodiment, when the classification result of the patent document group of the analyzed company is different from the classification result of the patent document group of the target company, the alliance candidate determination unit 36 determines the target company as the partner candidate company. Specifically, the alliance candidate determination unit 36 determines that the analysis result of the patent document group is different from the analysis result of the analyzed company among the plurality of target companies specified by the target company specifying unit 26. The selected target company is determined as a partner candidate company.

  In the first embodiment, the dissimilarity determination unit 34 has different issues in the focused technical field of the analyzed company and the focused technical field of the specific target company, but the elemental technologies are the same (pattern 1 in FIG. 1). It is determined that the classification result of the analyzed company is different from the classification result of the specific target company. In this case, the affiliation candidate determination unit 36 determines a specific target company as an affiliation candidate of the analyzed company.

  In addition, the difference determination unit 34, when the subject technical field of the analyzed company and the focused technical field of the specific target company have the same problem but different elemental technologies (pattern 2 in FIG. 1), the analyzed company It is determined that the classification result of the target company is different from the classification result of the specific target company. Also in this case, the affiliation candidate determination unit 36 determines a specific target company as an affiliation candidate of the analyzed company.

  In addition, the difference determination unit 34 categorizes the analyzed company when the focused technical field of the analyzed company and the focused technical field of the specific target company have different issues and different elemental technologies (pattern 3 in FIG. 1). It is determined that the classification results of a specific target company are different. Also in this case, the affiliation candidate determination unit 36 determines a specific target company as an affiliation candidate of the analyzed company.

  On the other hand, the difference determination unit 34 determines whether the analysis target company's focused technology field and the specific target company's focused technology field have the same problem and the same elemental technology (pattern 4 in FIG. 1). It is determined that the classification result and the classification result of the specific target company are the same. In this case, the alliance candidate determination unit 36 excludes the specific target company from the alliance candidates of the analyzed company.

  The technology range specifying unit 38 specifies the spread of the technical field of each company based on the patent document group of each of the analyzed company and the plurality of target companies. The technical range specifying unit 38 may specify the spread of the technical field of each company using a known technique for mapping each of a plurality of patent documents to any of a plurality of technical fields. For example, the technical scope specifying unit 38 uses the mapping analysis function (https://www.patent.ne.jp/service/macro/bunseki_mapping.html) provided by NIR Cyber Patent Co., Ltd. In addition, thermograph data indicating a technical area described in a plurality of patent documents may be generated. The technical range specifying unit 38 may count the number of technical fields related to patent documents of each company based on the thermograph data of each company, and specify the number as the spread of the technical field.

  As another method, the technical range specifying unit 38 determines the number of elemental technologies (or combinations of elemental technologies and issues) in which one or more patent documents exist based on the classification results of the patent documents by the document classification unit 32. You may count. And the number may be specified as the spread of the technical field.

  The rank determining unit 40 determines the rank of each partner candidate company for one or more target companies (hereinafter also referred to as “partner candidate companies”) as partner candidates determined by the partner candidate determining unit 36. The rank can be said to be a relative rank indicating the priority (in other words, the recommended level) as the partner.

  Analyzed companies (SMEs) are often at a disadvantage to the target companies (large companies). However, if the spread of the technical field handled by the analyzed company is larger than the spread of the technical field handled by the target company, it becomes easier for the analyzed company to advantageously proceed with the alliance / negotiation with the target company. Therefore, the rank determination unit 40 assigns higher ranks to partner candidates that have a smaller technical field. In addition, the rank determination unit 40 assigns a higher rank to a partner candidate company with a smaller technical field spread among the remaining partner candidate companies after focusing on a partner candidate company with a smaller technical field spread than the analyzed company. Also good.

  Further, the rank determination unit 40 assigns a higher rank to one or more partner candidate companies determined by the partner candidate determining unit 36 as the partner candidate company has a new application date for patent documents having similar contents to the analyzed company. . The patent document having contents similar to the analyzed company may be a patent document in which the combination of the problem and the elemental technology matches the focused technical field of the analyzed company. However, target companies whose focus technology field is the company's focus technology field are excluded from the alliance candidates. As described above, the rank determination unit 40 determines the rank of each partner candidate company according to the spread of the technical field of each partner candidate company and the recent patent application trend of each partner candidate company.

  The affiliation candidate output unit 42 transmits data including information on one or more affiliation candidate companies determined by the affiliation candidate determination unit 36 (for example, web page data) to the user device 12. The information on the partner candidate companies includes the rank of each partner candidate company determined by the rank determination unit 40. Further, the information on the partner candidate companies may be information in which a plurality of partner candidate companies are arranged higher as the rank is higher.

  The affiliation candidate output unit 42 may transmit information of affiliation candidate companies to the user device 12 as a synchronous response to the search request transmitted from the user device 12. Further, the information on the partner candidate company and the rank may be stored in a predetermined storage area. When the partner candidate output unit 42 receives a predetermined request different from the search request from the user device 12, the partner candidate output unit 42 may transmit the partner candidate company and rank information stored in the storage area to the user device 12. .

  Further, the affiliation candidate output unit 42 of the embodiment transmits information on the affiliation candidate companies determined by each of the patterns 1 to 3 illustrated in FIG. 1 to the user device 12. For example, the affiliation candidate output unit 42 extracts information on the affiliation candidate companies extracted in pattern 1 (difference in task, elemental technology match) in FIG. 1, and tie-up extracted in pattern 2 (match in issue, difference in elemental technology) in FIG. Information on candidate companies and information on partner candidate companies extracted in pattern 3 (differences in issues and differences in elemental technology) in FIG. 1 may be transmitted to the user device 12 in a distinguishable manner.

  As a modified example, in the search request transmitted from the user device 12, at least one of the patterns 1 to 3 in FIG. In this case, the affiliation candidate determination unit 36 may determine a target company that matches the pattern specified in the search request as an affiliation candidate company. The affiliation candidate output unit 42 may transmit information on the affiliation candidate companies determined based on the pattern specified in the search request to the user device 12.

  Further, the alliance candidate output unit 42 expresses the classification result of the patent document group possessed by each of the analyzed company and the alliance candidate company as information on the alliance candidate company in a two-dimensional chart centered on the subject and the elemental technology. The data (for example, the cross analysis result 50 in FIG. 4) may be transmitted to the user device 12. For example, the affiliation candidate output unit 42 may transmit data indicating the cross analysis result 50 of the analyzed company and the cross analysis result 50 of the affiliation candidate company side by side to the user device 12.

  In this case, it is desirable that the document classification unit 32 adds the document number symbol 52 at least in the focused technical field in the cross analysis result 50. In addition, when the document number symbol 52 is added to all the intersections between the problem and the elemental technology in the cross analysis result 50, the document classification unit 32 adds the document number symbol 52 having a different aspect from the other intersections to the focused technical field. It is desirable. For example, the document number symbol 52 having a different size and color from other intersections may be added, and the document number symbol 52 may be added in a manner that is more conspicuous than the symbols at other intersection points. Thereby, it is possible to easily show the user the reason for selecting the partner candidate company.

  In addition, when the document number symbol 52 is added to the intersection of the problem and the element technology in the cross analysis result 50, the document classification unit 32 or the partner candidate output unit 42 has one or more patents corresponding to the combination of the problem and the element technology. You may set the hyperlink which embedded the positional information on the literature (namely, the patent literature counted as a component of the intersection in which the literature number symbol 52 was installed). Thus, the user can easily browse patent documents corresponding to the problem and element technology indicated by the document number symbol 52 by selecting the specific document number symbol 52 of the cross analysis result 50.

The operation of the corporate alliance support system 10 having the above configuration will be described.
The user device 12 transmits a search request including identification information of the analyzed company to the partner search device 16 in accordance with a user operation. The document acquisition unit 22 of the partner search device 16 acquires the patent document group of the analyzed company designated by the search request from the technical document DB 14. The population specifying unit 24 of the partner search device 16 acquires a plurality of patent documents of other companies similar to each of the patent documents of the analyzed company from the technical document DB 14, and uses the acquired plurality of patent documents as the analysis population. Identify. The target company specifying unit 26 of the partner search device 16 analyzes large companies (for example, top five companies having high similarity) that are applicants or right holders of patent documents having high similarity to the patent documents of the analyzed company. Identify as target company.

  The problem extraction unit 28 of the partner search device 16 extracts problems to be solved described in each patent document for each company with respect to the analyzed company and the plurality of target companies. The technology extraction unit 30 of the partner search device 16 extracts elemental technologies that solve the problems described in each patent document for each company with respect to the analyzed company and the plurality of target companies. The document classification unit 32 of the partner search device 16 classifies the patent document group based on the combination of the problem and the element technology for each company with respect to the analyzed company and the plurality of target companies. For example, the cross analysis result 50 of FIG. 4 is generated for each company. The difference determination unit 34 of the partner search device 16 determines the difference in the classification result for each combination of the analyzed company and the target company. For example, it is determined which of patterns 1 to 4 in FIG. 1 corresponds to each combination of the analyzed company and the target company.

  The affiliation candidate determination unit 36 of the affiliation partner search device 16 determines a target company whose classification result by the document classification unit 32 is different from the analyzed company as the affiliation candidate company. For example, a target company corresponding to any one of patterns 1 to 3 in FIG. 1 is determined as a partner candidate company. The partner search device 16 may store the number of presentation candidates (for example, five companies) that is the number of partner candidate companies to be presented to the user. When the number of partner candidate companies determined by the partner candidate determination unit 36 is less than the number of candidates for presentation, the target company specifying unit 26 may specify a new target company. Thereafter, the partner candidate companies based on the new target company This determination process may be repeated.

  The technical range specifying unit 38 specifies the spread of the technical fields of the analyzed company and the plurality of target companies by a known mapping analysis. The rank determination unit 40 assigns a higher rank to a partner candidate company having a smaller technical field. The affiliation candidate output unit 42 transmits the affiliation candidate company and rank information to the user device 12. The user device 12 displays the information on the partner candidate companies and ranks transmitted from the partner search device 16 on the display and presents them to the user.

  According to the partner search device 16 of the first embodiment, it is possible to automatically search for partner candidates of the organization, and improve the efficiency and accuracy of the search. For example, the partner search device 16 automatically analyzes the technology possessed by each company by classifying the technical documents related to each of the small and medium enterprises and the plurality of large enterprises based on the issues and the axis of the elemental technology. Then, a large company having a technical complementary relationship for a small and medium-sized company can be presented to the user as an alliance candidate.

(Second embodiment)
In the first example, the difference between the classification result of the patent document group of the analyzed company and the classification result of the patent document group of the target company was determined by the difference in the focus technical field of both companies. In the second embodiment, for each company, the scores in the patterns 1 to 4 shown in FIGS. 1A to 1D are derived based on the classification result of the patent document group, and the classification result is based on the score. Judge the difference between.

  The configuration of the corporate alliance support system 10 of the second embodiment is the same as the configuration of the corporate alliance support system 10 of the first embodiment (FIG. 2). The configuration of the partner search device 16 of the second embodiment is also the same as the configuration of the partner search device 16 of the first embodiment (FIG. 3). Hereinafter, the description of the contents already described in the first embodiment will be omitted, and differences from the first embodiment will be mainly described.

  Here, the classification result of the patent document group of the analyzed company by the document classification unit 32 (for example, the data of the cross analysis result 50 in FIG. 4) and the classification result of the patent document group of the target company by the document classification unit 32 have been generated. To do. The difference determination unit 34 derives the scores in the patterns 1 to 4 shown in FIGS. 1A to 1D based on the classification result of each company. Hereinafter, each pattern will be described.

A method for deriving the score of pattern 1 (difference between tasks, element technology match) will be described.
FIGS. 5A to 5D and FIGS. 6A to 6D show examples of derivation of the pattern 1 score. 5A shows the classification result of the patent document group of the company X (analyzed company), and FIG. 5B shows the classification result of the patent document group of the company Y (one target company). . For example, FIG. 5A shows that in the group of patent documents of company X, there are five documents in which the problem A is solved using the technique 1.

  The difference determination unit 34 adds the number of documents for each elemental technology of company X (FIG. 5C), and adds the number of documents for each elemental technology of company Y (FIG. 5D). The difference determination unit 34 compares the total number of cases for each company, and identifies elemental technologies for which both companies have an application. In the example of FIG. 5C and FIG. 5D, the technique 1 and the technique 3 are specified. The difference determination unit 34 changes the component (the number of documents) to 0 for technologies other than the identified technologies (Technology 1 and Technology 3). FIG. 6A shows that the components of “issue A” and “technology 2” of company X have been changed to 0, and FIG. 6B shows the components of “issue C” and “technology 4” of company Y. Is changed to 0.

  The difference determination unit 34 subtracts each component of the company Y in FIG. 6B from each component of the company X in FIG. 6A (FIG. 6C). The difference determination unit 34 converts each component after subtraction into an absolute value (FIG. 6D). The difference determination unit 34 adds the absolute values of the components, and uses the combined value as the score of the pattern 1. In the example of FIG. 6D, the score is “9”. The difference determination unit 34 determines that the company X and the company Y match the pattern 1 when the score of the pattern 1 is equal to or greater than a predetermined threshold. The difference determination unit 34 determines whether or not the combination of the analyzed company and the target company matches the pattern 1.

The score derivation method for pattern 2 (issue matching, element technology difference) will be described.
7 (a)-(d) and FIGS. 8 (a)-(b) show examples of derivation of the score of pattern 2. FIG. 7A shows the classification result of the patent document group of the company X (analyzed company), and FIG. 7B shows the classification result of the patent document group of the company Y (one target company). .

  The difference determination unit 34 adds the number of documents for each task of the company X (FIG. 7C), and adds the number of documents for each task of the company Y (FIG. 7D). The difference determination unit 34 compares the total number of cases for each company, and identifies an issue for which both companies have an application. In the example of FIG. 7C and FIG. 7D, the task A is specified. The difference determination unit 34 changes the component (the number of documents) to 0 for tasks other than the identified task (task A). Company X has no change in its components from FIG. On the other hand, the components of “Problem B” and “Technology 3” and “Problem C” and “Technology 4” of Company Y are changed to 0 (FIG. 8A).

  The difference determination unit 34 subtracts each component of the company Y (FIG. 8A) from each component of the company X (FIG. 7A) (FIG. 8B). The difference determination unit 34 converts each component after subtraction into an absolute value. The difference determination unit 34 adds the absolute values of the components, and uses the combined value as the score of the pattern 2. In the example of FIG. 8B, the score is “10”. The difference determination unit 34 determines that the company X and the company Y match the pattern 2 when the score of the pattern 2 is equal to or greater than a predetermined threshold. The difference determination unit 34 determines whether or not the combination of the analyzed company and the target company matches the pattern 2.

A score derivation method for pattern 3 (difference in task, difference in element technology) will be described.
FIGS. 9A to 9D show examples of derivation of the pattern 3 score. FIG. 9A shows the classification result of the patent document group of the company X (analyzed company), and FIG. 9B shows the classification result of the patent document group of the company Y (one target company). .

  The difference determination unit 34 specifies a combination of a problem and a technology whose number of documents of the company X is zero. In the example of FIG. 9A, “Problem B”, “Technology 4”, “Problem B”, “Technology 5”, “Problem C”, “Technology 4”, “Problem C”, “Technology 5”, “Problem D”, “ A combination of “Technology 4”, “Issue D”, “Technology 5”, “Issue E”, “Technology 4”, “Issue E”, “Technology 5” is specified. The difference determination unit 34 changes components other than the specified combination to 0 among the components of the classification result of the company Y. In the example of FIG. 9B, the components of “Problem A” and “Technology 1” and the components of “Problem B” and “Technology 3” are changed to 0 (FIG. 9C).

  The difference determination unit 34 identifies a combination of a problem and a technique in which the number of documents of the company Y is zero. In the example of FIG. 9B, a combination of “Problem D”, “Technology 2”, “Problem D”, “Technology 5”, “Problem E”, “Technology 2”, “Problem E”, “Technology 5” is specified. The difference determination unit 34 changes components other than the identified combination to 0 among the components of the classification result of the company X. In the example of FIG. 9A, the components of “Problem A” and “Technology 1”, “Problem A” and “Technology 2”, and “Problem A” and “Technology 3” are changed to 0 (FIG. 9). 9 (d)).

  The difference determination unit 34 adds each component of the company X after the component adjustment (FIG. 9D) and each component of the company Y after the component adjustment (FIG. 9C). Score. In the example of FIG. 9C and FIG. 9D, the score is “2”. The difference determination unit 34 determines that the company X and the company Y match the pattern 3 when the score of the pattern 3 is equal to or greater than a predetermined threshold. The difference determination unit 34 determines whether or not the combination of the analyzed company and the target company matches the pattern 3.

The score derivation method of pattern 4 (issue matching, element technology matching) will be described.
FIGS. 10A to 10C show examples of the derivation of the pattern 4 score. FIG. 10A shows the classification result (ID of each component) of the patent document group of company X (analyzed company), and FIG. 10B shows the patent document of company Y (one target company). The group classification results are shown.

  The difference determination unit 34 adds the products of the component values of the same combination of the task and the elemental technology, and uses the total value as the score of the pattern 4 as represented by Expression 1 in FIG. The difference determination unit 34 determines that the company X and the company Y match the pattern 4 when the score of the pattern 4 is equal to or greater than a predetermined threshold. The difference determination unit 34 determines whether or not the combination of the analyzed company and the target company matches the pattern 4.

  If the analyzed company and the target company match at least one of Pattern 1, Pattern 2, and Pattern 3 and do not match Pattern 4, the partner candidate determination unit 36 determines the target company as the partner candidate company. In other words, when the analyzed company and the target company match pattern 4, the alliance candidate determination unit 36 forms a partnership with the target company regardless of whether it matches at least one of pattern 1, pattern 2, and pattern 3. Exclude from candidate companies.

  The affiliation candidate output unit 42 may transmit, as information about the affiliation candidate company, information on a plurality of affiliation candidate companies arranged higher as the pattern score is higher to the user device 12 for each pattern. Further, the affiliation candidate output unit 42 transmits, to the user apparatus 12, information on a plurality of affiliation candidate companies arranged higher as the pattern score is higher and the rank is higher, as information on the affiliation candidate companies. May be.

  In the partner search apparatus 16 of the second embodiment, as in the first embodiment, it is possible to automatically search for partner candidates for the organization and improve the efficiency and accuracy of the search. Further, in the second embodiment, since the difference of the classification result is determined based on a plurality of components in the classification result of the patent document of each company, the accuracy of the determination of the difference can be improved.

  The present invention has been described based on the first and second embodiments. These embodiments are exemplifications, and it is understood by those skilled in the art that various modifications can be made to the combination of each component and each processing process, and such modifications are within the scope of the present invention. is there.

  A first modification will be described. In the calculation of the score of pattern 1 in the second example, the components other than the technology for which both companies were applied were changed to 0, and then the components of the classification results of other companies were subtracted from the components of the classification results of one company. As a modified example, addition may be performed instead of subtraction.

  A second modification will be described. In the said 2nd Example, all the scores of the patterns 1-4 were calculated in the number of patent documents. As a modified example, the score of each pattern may be calculated after converting the number of patent documents into a ratio (normalization conversion). Generally, SMEs (analyzed companies) have fewer patent applications than large companies, but according to this modification, patent documents related to relatively few analyzed companies are easily reflected in the score.

  A score derivation method for pattern 1 (difference between tasks, element technology match) in this modification will be described. 11 (a)-(d) and FIGS. 12 (a)-(d) show examples of derivation of the score of pattern 1 in the second modification. Although not shown, the classification result of the patent document group of the company X (analyzed company) is as shown in FIG. 5A, and the classification result of the patent document group of the company Y (one target company) is shown in FIG. As shown in b).

  The difference determination unit 34 converts the component of the classification result of the company X from the number of patent documents into a ratio (FIG. 11A), and converts the component of the classification result of the company Y from the number of patent documents into a ratio (FIG. 11). 11 (b)). Note that the document classification unit 32 outputs a classification result (for example, FIG. 11A and FIG. 11B) based on the ratio of patent documents instead of the classification result based on the number of patent documents. Also good. The difference determination unit 34 adds the document ratios for each elemental technology of the company X (FIG. 11C), and adds the document ratios for each elemental technology of the company Y (FIG. 11D). The difference determination unit 34 compares the summing ratios for each company, and identifies the elemental technologies for which both companies have an application. In the example of FIG. 5C and FIG. 5D, the technique 1 and the technique 3 are specified.

  The difference determination unit 34 changes the component (document ratio) to 0 for technologies other than the identified technologies (Technology 1 and Technology 3). 12A shows that the components of “issue A” and “technology 2” of company X are changed to 0, and in FIG. 12B, the components of “issue C” and “technology 4” of company Y are shown. Is changed to 0. The difference determination unit 34 subtracts each component of the company Y in FIG. 12B from each component of the company X in FIG. 12A (FIG. 12C). The difference determination unit 34 converts each component after subtraction into an absolute value (FIG. 12D).

  The difference determination unit 34 adds the absolute values of the components, and uses the combined value as the score of the pattern 1. In the example of FIG. 12D, the score is “81”. The difference determination unit 34 determines that the company X and the company Y match the pattern 1 when the score of the pattern 1 is equal to or greater than a predetermined threshold. Similar to the pattern 1, the difference determination unit 34 may calculate the scores of the patterns 2 to 4 with the patent document ratio as a unit.

  A third modification will be described. In the above-described example, the difference in the classification results of the patent documents was determined between the two companies of the analyzed company and the target company. As a modified example, the difference in the classification result of the patent document may be determined between three or more companies by the analyzed company and two or more target companies. In other words, the partner search device 16 may propose partnerships with a plurality of companies to the analyzed company.

  A fourth modification will be described. The search request transmitted from the user device 12 may include a partner candidate condition designated by the user. The condition of the alliance candidate may be an industry, a business type, a business type, or a financial condition such as capital. The partner search device 16 may further include a company information acquisition unit that acquires outline information, financial information, and the like of the company from an external company information DB (not shown). The target company specifying unit 26 may narrow down the target companies for analysis so as to match the conditions of the tie-up candidates indicated by the search request based on the information acquired by the company information acquisition unit. For example, the target company specifying unit 26 specifies, as a target company, a company that is associated with a patent document having a high degree of similarity with the patent document of the analyzed company and that matches the partner candidate condition indicated by the search request. May be.

  A fifth modification will be described. In the above embodiment, issues and technologies are extracted from one or more types of documents such as patent documents, academic papers, technical manuals, product manuals, white papers, etc., and determined as tie-up candidates. However, in the corporate tie-up support system of the fifth embodiment, the document type to be processed is identified, the issues and technologies are extracted and classified for the document type specified by the user, and determined as a tie-up candidate. Also good. In addition, issues and techniques can be extracted and classified for each document type or a document for each combination of document types, and each can be determined as a partner candidate. By doing so, for example, it should be affiliated with Company A from the viewpoint of patent literature, should be affiliated with Company B from the viewpoint of academic papers, and affiliated with Company C from the viewpoint of technical explanation It is possible to obtain an output result indicating that it should be affiliated with Company A from the viewpoint of the product manual, and that it should be affiliated with Company C from the viewpoint of the white paper. Therefore, it is possible to obtain an output result indicating that the company should tie up with the company A and should tie up with the company C from the viewpoint of the technical manual and the product manual. Thus, by obtaining different output results depending on the document type and comparing them, it becomes possible to analyze a more appropriate partner.

  Any combination of the above-described embodiments and modifications is also useful as an embodiment of the present invention. The new embodiment resulting from the combination has the effects of the combined example and modification. Further, it should be understood by those skilled in the art that the functions to be fulfilled by the constituent elements described in the claims are realized by the individual constituent elements shown in the embodiments and the modified examples or by their cooperation.

  DESCRIPTION OF SYMBOLS 10 company tie-up support system, 16 tie-up partner search apparatus, 28 assignment extraction part, 30 technology extraction part, 32 document classification part, 34 difference judgment part, 36 tie-up candidate decision part, 42 tie-up candidate output part.

Claims (5)

An issue extraction unit that extracts issues described in a group of technical documents related to the organization for each of a plurality of organizations;
A technology extraction unit that extracts elemental technologies for solving the problem described in the technical document group for each of the plurality of organizations;
For each organization, a classification unit that classifies the technical document group based on a combination of the problem and the elemental technology;
A determination unit that determines at least two organizations with different classification results by the classification unit as affiliation candidates;
An output unit for outputting information on the alliance candidate determined by the determination unit;
An information processing apparatus comprising:   The output unit outputs data representing a classification result of a technical document group for each organization as a two-dimensional chart with an issue and an elemental technology as an axis as information on the tie-up candidate. The information processing apparatus described in 1.   The said determination part determines the at least 2 organization which became the classification result which shows that the subject as described in the said technical document group is the same, but elemental technologies differ, as a tie-up candidate, It is characterized by the above-mentioned. The information processing apparatus described in 1.   The said determination part determines the at least 2 organization which became the classification | category result which shows that the elemental technologies as described in the said technical document group are the same, but a subject is different, as a tie-up candidate. The information processing apparatus according to any one of the above. A function for extracting issues described in technical documents related to the organization for each of a plurality of organizations;
A function of extracting elemental technologies for solving the problem described in the technical document group for each of the plurality of organizations;
For each organization, a function of classifying the technical document group based on a combination of the problem and the elemental technology,
A function of determining at least two organizations with different classification results according to the function to be classified as affiliation candidates;
A function for outputting information on the alliance candidate determined by the function to be determined;
A computer program for realizing a computer.

Images