JP7586184B2 - BUSINESS MATCHING SYSTEM, BUSINESS MATCHING METHOD, AND BUSINESS MATCHING PROGRAM - Google Patents

Description

The present invention relates to business matching between companies.

When conducting transactions between companies, a business matching system is used that presents combinations of companies suitable for the transaction. The business matching system predicts companies that are suitable as trading partners for the companies to be matched, and proposes candidate companies for matching based on the prediction results. For example, technology for performing such matching is disclosed in Patent Document 1.

Patent Document 1 relates to a business matching system that suggests candidate companies for matching based on data including past transaction data. The business matching system of Patent Document 1 classifies multiple companies, including the target company for matching, into segments based on transaction performance data, etc. The business matching system of Patent Document 1 presents as candidate companies for matching companies that do not do business with the target company, from among companies of the same type as the target company for matching and that have high values for preset indicators.

JP 2017-182243 A

The business matching system in Patent Document 1 predicts companies that would be suitable business partners based on data including past transaction data. This makes it difficult to make proposals that reflect the needs of the target companies.

In order to solve the above problems, the present invention aims to provide a business matching system etc. that can present potential business partners that meet the needs of the target company regardless of whether or not transaction data is available.

In order to solve the above problems, the business matching system of the present invention is a business matching system that matches businesses between companies and includes a matching unit. The matching unit outputs a matching score between the first company and the second company based on the similarity between a natural language sentence that indicates the needs of the first company and company information including the industry of the second company and product information related to products handled by the second company.

The business matching method of the present invention is a business matching method for matching businesses between companies, which outputs a matching score between a first company and a second company based on the similarity between a natural language sentence indicating the needs of the first company and company information including the industry of the second company and product information regarding products handled by the second company.

The program recording medium of the present invention is a program recording medium having recorded thereon a business matching program for matching business between companies. The business matching program causes a computer to execute a process for outputting a matching score between a first company and a second company based on the degree of similarity between a natural language sentence indicating the needs of a first company and company information including the industry of the second company and product information regarding products handled by the second company.

According to the present invention, it is possible to present potential business partners that meet the needs of the target company, regardless of whether or not transaction data is available.

1 is a diagram showing an outline of a configuration of a first embodiment of the present invention; FIG. 2 is a diagram showing an example of a needs statement according to the first embodiment of the present invention. FIG. 2 is a diagram showing an example of a sentence responding to a needs sentence in the first embodiment of the present invention. FIG. 2 is a diagram illustrating a configuration of an extraction unit according to the first exemplary embodiment of the present invention. FIG. 4 is a diagram illustrating a process in an extraction unit according to the first embodiment of the present invention. FIG. 2 is a diagram illustrating a configuration of a matching unit according to the first exemplary embodiment of the present invention. FIG. 4 is a diagram illustrating a process in a business distance calculation unit according to the first embodiment of the present invention. FIG. 2 is a diagram showing an operation flow of the business matching system according to the first embodiment of the present invention. FIG. 2 is a diagram illustrating a process of extracting a phrase according to the first embodiment of the present invention. FIG. 2 is a diagram illustrating a process for determining a business expression according to the first embodiment of the present invention. FIG. 2 is a diagram illustrating an example of a display screen according to the first embodiment of the present invention. FIG. 13 is a diagram showing an outline of the configuration of a second embodiment of the present invention. A diagram illustrating the configuration of a negotiation success/failure score calculation unit in the second embodiment of the present invention. FIG. 13 is a diagram illustrating a process for calculating a negotiation success/failure score according to the second embodiment of the present invention. FIG. 11 is a diagram illustrating a configuration of a matching unit according to a second exemplary embodiment of the present invention. FIG. 11 is a diagram showing an operation flow of the business matching system according to the second embodiment of the present invention. FIG. 11 is a diagram illustrating an example of a display screen according to the second embodiment of the present invention. FIG. 11 is a diagram illustrating an example of a display screen according to the second embodiment of the present invention. FIG. 13 is a diagram illustrating an example of another configuration of the second exemplary embodiment of the present invention. FIG. 13 is a diagram showing an outline of the configuration of a third embodiment of the present invention. FIG. 11 is a diagram showing an operation flow of the business matching system according to the third embodiment of the present invention. FIG. 13 is a diagram showing an example of another configuration of the present invention.

(First embodiment)
A first embodiment of the present invention will be described in detail with reference to the drawings. Fig. 1 is a diagram showing an outline of the configuration of a business matching system 10 of this embodiment. The business matching system 10 of this embodiment includes an acquisition unit 11, an extraction unit 12, an importance calculation unit 13, a matching unit 14, and an output unit 15.

The acquisition unit 11, extraction unit 12, importance calculation unit 13, matching unit 14 and output unit 15 of the business matching system 10 may be provided on the same server, or may be distributed and provided on different servers and connected via a network.

The business matching system 10 of this embodiment is a system that presents candidates for companies suitable as trading partners when a company does business with another company. Specifically, when the target company for matching is a first company and one or more candidate companies that are candidates for the matching destination are the second companies, the business matching system 10 is a system that calculates the similarity between the needs statement of the first company and the company information of the second company, and presents the matching destination company based on the calculated similarity. The company information of the second company, i.e., the candidate company, includes product information handled by the second company. For example, the product information may be any of the following information: the product name of the product handled by the candidate company, the service name of the service provided by the candidate company, or the system name of the system provided by the candidate company.

Here, a company suitable as a trading partner, that is, a candidate company that can be a matching partner, refers to, for example, a candidate trading partner that can supply products in the field of a new business when a matching target company starts a new business. Also, a matching target company refers to a company that searches for a matching partner when conducting business matching. Also, candidate companies that are suitable as trading partners for a matching target company are also called matching candidate companies.

The configuration of the business matching system 10 will be explained with reference to Figure 1.

The acquisition unit 11 acquires the needs text of the target company for matching and the company information of the candidate companies for matching. The needs text is a natural language text that describes information including the contents of the transaction and the business details when the target company for matching seeks a business partner. Figure 2 shows an example of a needs text. In the needs text of Figure 2, the contents of the business to be implemented and the requests to the business partner are described in natural language text.

The corporate information of a candidate company for matching refers to, for example, information on one or more of the following items: the candidate company's industry, client companies, sales, profits, number of employees, factory location, distribution base, business details such as products handled, past products sold, and services provided, and intellectual property rights ownership status. The corporate information of a candidate company for matching may be information other than the items exemplified above. In addition, the corporate information of a candidate company for matching may include natural language text when the candidate company for matching responds to a needs text. Figure 3 shows an example of a natural language text in which a candidate company for matching responds to a needs text transmitted by a target company. In Figure 3, information including the business details and products that the candidate company can provide that are thought to match the needs text is written in natural language text.

The acquisition unit 11 also acquires business data that serves as a reference when the extraction unit 12 extracts phrases from the needs text and company information. Hereinafter, the business data that serves as a reference when the extraction unit 12 extracts phrases is referred to as reference business data. As the reference business data, for example, data from the Japan Standard Industrial Classification can be used. The reference business data may be data other than the Japan Standard Industrial Classification as long as it contains business terms in the field to which the target company for matching belongs. The acquisition unit 11 acquires the reference business data from, for example, a data providing server of a government agency. The reference business data may be stored in advance in the matching system 10 for each field to which the company belongs.

The extraction unit 12 extracts phrases that are important for matching from the needs text and company information. Figure 4 shows the configuration of the extraction unit 12. The extraction unit 12 further includes a phrase extraction unit 21 and a business expression determination unit 22.

The phrase extraction unit 21 extracts phrases from reference business data, needs sentences and company information.

The phrase extraction unit 21 extracts phrases and calculates the feature amounts of each phrase based on the reference business data, the needs text, and the company information, for example, using the Bidirectional Encoder Representations from Transformers (BERT) method. The phrase extraction unit 21 may also extract phrases and calculate the feature amounts using other methods such as Word2Vec.

The phrase extraction unit 21 may extract phrases using a learning model generated by machine learning. The learning model used to extract phrases is generated by machine learning using deep learning with a corpus as input data. The phrase extraction unit 21 extracts phrases by using the learning model generated in advance using the corpus as a named entity extractor.

The learning model used to extract phrases may be generated according to the business fields of the target and candidate companies. In this case, the phrase extraction unit 21 generates a first learning model for extracting phrases by deep learning using teacher data in which sample sentence data and sample company information are used as learning input data and phrases extracted from the reference business data are used as positive example label data. Next, phrases are extracted from the needs sentence and company information using the first learning model, and a first feature amount is calculated for the extracted phrases. The phrase extraction unit 21 also generates a second learning model for extracting phrases of business expressions from the sample business data by deep learning. Here, the business expressions refer to language expressions used between the target and candidate companies based on a common understanding in the field in which the target and candidate companies conduct transactions. The business expressions refer to words and phrases used when communicating between companies, such as, for example, creating needs sentences, negotiating, creating contracts, and creating technical documents. The phrase extraction unit 21 calculates a second feature amount for the phrases of business expressions extracted from the reference business data using the second learning model. It is assumed that the calculation algorithm for the first feature amount and the second feature amount is the same.

The business expression determination unit 22 uses the phrase extracted from the reference business data as a positive example to determine whether the phrase extracted from the needs text and the company information is a business expression. When the above-mentioned first learning model and second learning model are used, the business expression determination unit 22 uses the phrase extracted from the reference business data by the second learning model as a positive example to determine whether the phrase extracted from the needs text and the company information by the first learning model is a business expression. Specifically, the business expression determination unit 22 determines whether each phrase extracted from the needs text and the company information and characterized is a business expression based on the similarity with a business expression. When determining whether the phrase extracted from the needs text and the company information is a business expression, the business expression determination unit 22 calculates the probability that the phrase is a business expression. The business expression determination unit 22 determines that the extracted phrase is a business expression when the probability that the extracted phrase is a business expression exceeds a preset threshold.

Figure 5 is a diagram showing the processing contents of the phrase extraction unit 21 and the business expression determination unit 22. In the upper part of Figure 5, the phrase extraction unit using the second learning model extracts business expression phrases from the reference business data and converts them into features. In the upper part of Figure 5, the phrase extraction unit using the first learning model extracts phrases from the needs text and company information and converts them into features. In the lower part of Figure 5, the business expression determination unit classifies each phrase that has been converted into features into business expressions and non-business expressions, and extracts business expression phrases. Note that, although an example of extracting phrases using two different learning models has been described above, phrases may be extracted using one common learning model. In other words, the number of learning models used to extract phrases is not limited to one or two, and learning models may be generated for each needs text, each target company, each candidate company, etc.

The importance calculation unit 13 calculates the importance of the phrases extracted by the extraction unit 12. The importance of a phrase refers to an index of the importance of the phrase in transactions between the matching target company and the candidate company. The importance calculation unit 13 calculates the importance of each phrase, for example, by the TF-IDF (Term Frequency - Inverse Document Frequency) method. The importance calculation unit 13 may also calculate the importance of each phrase using other methods such as the Okapi BM25 method.

The matching unit 14 calculates an index indicating the matching suitability of the candidate companies as a matching score. FIG. 6 is a diagram showing the configuration of the matching unit 14. The matching unit 14 further includes a business distance calculation unit 23.

The business distance calculation unit 23 calculates the closeness between the needs of the matching target company and the needs of the candidate company. The business distance calculation unit 23 weights each phrase based on its importance and adds them together to calculate the closeness between the needs of the matching target company and the needs of the candidate company, and outputs the result as a matching score.

The output unit 15 outputs the matching result in the matching unit 14. The output unit 15 may be a display control unit that controls a display device to display the matching result.

Figure 7 is a diagram showing the processing contents in the business distance calculation unit 23. The business distance calculation unit 23 calculates the similarity of the phrase extracted from the company information and characterized for each phrase by brute force, with respect to the phrase extracted from the needs text and characterized. For example, the business distance calculation unit 23 calculates the similarity of the phrase as a cosine similarity. The business distance calculation unit 23 may calculate the similarity of the phrase as a Euclidean distance between the feature vectors. Note that the similarity is not limited to the above cosine similarity and the distance between the feature vectors. The business distance calculation unit 23 adds the phrases with the highest similarity for each phrase, weighting them according to their importance, and calculates the result as a business score. The business distance calculation unit 23 calculates a business score for each candidate company.

The operation of the business matching system 10 of this embodiment will be described. Figure 8 is a diagram showing the operation flow of the business matching system 10.

The acquisition unit 11 acquires reference business data that is used as a reference when extracting phrases. The acquisition unit 11 also acquires the needs statements of the companies to be matched and the company information of the candidate companies to be matched (step S11).

Once the needs sentence and company information are acquired, the phrase extraction unit 21 of the extraction unit 12 extracts phrases from the business data, the needs sentence and the company information (step S12). The phrase extraction unit 21 extracts phrases from the business data, the needs sentence and the company information using, for example, the BERT method, and converts them into features. Figure 9 is a diagram that illustrates the process of extracting phrases from needs sentences. In Figure 9, a needs sentence similar to that in Figure 2 is input to the phrase extraction unit 21. The phrase extraction unit 21 extracts, for example, "agricultural support" as a phrase from the needs sentence.

Once a phrase is extracted, the business expression determination unit 22 uses the phrase extracted from the business data as a positive example to determine whether the phrase extracted from the needs sentence and the company information is a business expression. If the extracted phrase is not a business expression (No in step S13), the process returns to step S12, and the phrase extraction unit performs an operation to extract other phrases that have not yet been extracted from the needs sentence and the company information.

For example, when a phrase extracted from the needs text and company information does not exist in the positive examples based on business expressions, the business expression determination unit 22 calculates the probability that the phrase is a business expression. When the probability that the extracted phrase is a business expression exceeds a preset threshold, the business expression determination unit 22 determines that the extracted phrase is a business expression.

FIG. 10 is a diagram showing a schematic diagram of a process for determining whether an extracted phrase is a business expression. In FIG. 10, data of the Japanese Standard Industrial Classification is used as the reference business data. In the example of FIG. 10, the phrase "AI technology" included in the needs sentence does not exist in the Japanese Standard Industrial Classification. At this time, for example, the business expression determination unit 22 calculates the probability that "AI technology" is a business expression based on the feature amount, using the phrase extracted from the reference business data as a positive example. The business expression determination unit 22 calculates the probability that "AI technology" is a business expression, for example, using a PU Learning (Positive Unlabeled Learning) method. After calculating the probability that it is a business expression, the business expression determination unit 22 compares the calculated probability with a threshold value. At this time, for example, if the probability that "AI technology" is a business expression is 0.87 and the determination threshold is 0.75, the business expression determination unit 22 determines that "AI technology" is a business expression. In addition, the business expression determination unit 22 may calculate the similarity of "AI technology" to a phrase extracted from reference business data based on features, and compare the calculated similarity with a threshold value, regarding it as a probability indicating whether it is a business expression.

If the phrase extracted in step S12 is a business expression (Yes in step S13), the extraction unit 12 outputs the phrase determined to be a business expression to the importance calculation unit 13.

Upon receiving a phrase determined to be a business expression, the importance calculation unit 13 calculates the importance of each phrase in business matching (step S14). The importance calculation unit 13 calculates the importance of each phrase, for example, by the TF-IDF method. The importance is used to weight each phrase based on its importance in business matching. A needs sentence contains a mixture of phrases that are highly important in matching and phrases that are less important. Furthermore, the importance of each phrase in matching may differ depending on the field and needs in which matching is performed. Therefore, by using the importance of each phrase in matching, it is possible to remove noise and improve the accuracy of matching.

Once the importance of each phrase is calculated, the business distance calculation unit 23 of the matching unit 14 calculates the closeness between the needs of the matching target company and the needs of the candidate companies as a matching score (step S15). For example, the business distance calculation unit 23 calculates the Euclidean distance between the needs of the target company and the needs of each of the candidate companies, and calculates the matching score by adding the similarity between the needs of the target company and the needs of each of the candidate companies weighted by the importance. The business distance calculation unit 23 may also calculate the matching score using only the similarity between the needs of the target company and the needs of each of the candidate companies without using the importance.

Once the matching score has been calculated, the output unit 15 outputs the matching result (step S16). The output unit 15 outputs the names and matching scores of a preset number of candidate companies, for example, in order of highest matching score. Figure 11 is a schematic diagram of an example of a display screen on which candidate companies for matching are output in order of highest matching score. In the example of Figure 11, the output unit 15 displays the names of candidate companies for matching in order of highest matching score, which is displayed as a score. Users of the business matching system 10 can refer to the display screen to obtain information on candidate companies suitable for trading.

The output unit 15 may, for example, display on the display device only the name of the candidate company with the highest matching score and the matching score. The output unit 15 may output the name of the candidate company, the matching score, and one or both of the response document and company information of the candidate company. The output unit 15 may also output the name of the candidate company whose similarity is equal to or exceeds a preset standard and the matching score. The output unit 15 may also output the name of the candidate company, the matching score, and one or both of the text used in the response by the candidate company and the company information. The output unit 15 may be a display control unit that controls the display device to display the information outputted above.

The business matching system 10 of this embodiment extracts phrases from the needs sentences, which are natural language sentences sent by the target companies for matching, and calculates the similarity between the phrases extracted from the company information of the candidate companies. The business matching system 10 of this embodiment also uses the similarity to calculate an index indicating the suitability of matching between the target companies for matching and the candidate companies as a matching score. The business matching system 10 of this embodiment also uses the importance of each phrase in addition to the similarity when calculating the matching score, thereby improving the accuracy of estimating candidate companies suitable for matching with the target company. The business matching system 10 of this embodiment also performs matching based on the needs sentences sent by companies recruiting trading partners and the company information of the candidate companies, so it can predict candidate companies suitable for matching regardless of the presence or absence of past trading data. In addition, since the prediction is made based on the needs sentences written in natural language, the user of the business matching system 10 can make predictions without preparing data in a format different from the sentences used in business, which is highly convenient.

Second Embodiment
The second embodiment of the present invention will be described in detail with reference to the drawings. Fig. 12 is a diagram showing an outline of the configuration of a business matching system 30 of this embodiment. The business matching system 30 of this embodiment includes an acquisition unit 11, an extraction unit 12, an importance calculation unit 13, a negotiation success/failure score calculation unit 31, a matching unit 32, and an output unit 15.

The acquisition unit 11, extraction unit 12, importance calculation unit 13, negotiation feasibility score calculation unit 31, matching unit 32 and output unit 15 of the business matching system 30 may be provided on the same server, or on multiple servers and connected via a network.

The business matching system 10 of the first embodiment calculates a matching score based on the similarity and importance of phrases extracted from the needs text and company information, and presents candidate companies. In addition to this configuration, the business matching system 30 of this embodiment further uses negotiation information including data on the feasibility of negotiations in similar past negotiations when calculating the matching score. This makes it possible to improve the accuracy of matching in this embodiment.

The configurations and functions of the acquisition unit 11, extraction unit 12, importance calculation unit 13 and output unit 15 of the business matching system 30 of this embodiment are similar to those of the same named parts of the business matching system 10 of the first embodiment.

The negotiation feasibility score calculation unit 31 calculates the negotiation success rate based on the similarity with past negotiation results. Figure 13 is a diagram showing the configuration of the negotiation feasibility score calculation unit 31. The negotiation feasibility score calculation unit 31 further includes a feature calculation unit 41 and a negotiation similarity calculation unit 42.

The feature calculation unit 41 converts each piece of information related to the target company and candidate companies into features. The feature calculation unit 41 extracts information on the type of needs, desired industry of matching, and industry from the needs text of the target company. The feature calculation unit 41 also extracts information on the desired industry of matching and industry from the company information of the candidate companies. The feature calculation unit 41 converts the extracted information into features using a feature calculator appropriate for each item.

The processing in the feature calculation unit 41 is performed using a learning model generated by deep learning using a neural network (NN) such as a convolutional neural network (CNN). The learning model uses information on the target company and candidate companies as input data, and is generated by machine learning using deep learning with track record data on whether negotiations are successful (whether negotiations are successful or not) as labels. Negotiation information may include transaction amounts, number of transactions, frequency of transactions, or duration of transactions in past negotiations. Negotiation information is not limited to the above, as long as it is information related to negotiations.

The negotiation similarity calculation unit 42 calculates the similarity with the past negotiation history based on the features calculated by the feature calculation unit 41, and calculates the negotiation feasibility score. The negotiation similarity calculation unit 42 may calculate the negotiation feasibility score as the product of the success rate of the negotiation in similar past negotiations and the similarity.

Figure 14 is a schematic diagram showing the calculation process of the negotiation feasibility score in the negotiation feasibility score calculation unit 31. In Figure 14, the negotiation feasibility score calculation unit extracts information on the type of needs "increased procurement", the desired industries of the matching candidates "manufacturing, trading company, consulting", and the target company's industry "wholesale" from the needs text of the target company for matching. In addition, the negotiation feasibility score calculation unit extracts information on the type of needs of the candidate company "expanding sales channels" and the industry "manufacturing" from the company information of the candidate company. The negotiation feasibility score calculation unit inputs each piece of extracted information into a feature calculator appropriate for each item and converts it into features.

Once the information for each item has been converted into features, the past negotiation similarity calculation unit calculates a negotiation feasibility score based on the similarity with the similarity of past negotiations. In the example of Figure 14, the negotiation feasibility score for company A is output as 0.629.

The matching unit 32 calculates a matching score, which is an index showing the suitability of the matching target companies as trading partners, based on the business score and the negotiation feasibility score. FIG. 15 is a diagram showing the configuration of the matching unit 32. The matching unit 32 further includes a business distance calculation unit 43 and a matching score calculation unit 44.

The business distance calculation unit 43 calculates the closeness between the needs of the matching target company and the needs of the candidate company as a business score. The business distance calculation unit 43 calculates the closeness between the needs of the matching target company and the needs of the candidate company as a business score in a manner similar to that of the business distance calculation unit 43 in the first embodiment. The business distance calculation unit 43 calculates the similarity between the needs of the target company and the information of the candidate company using, for example, cosine similarity to calculate the business score. The business distance calculation unit 43 may also calculate the similarity using other methods such as Euclidean distance to calculate the business score.
The matching score calculation unit 44 calculates a matching score based on the business score and the negotiation feasibility score. For example, the matching score calculation unit 44 calculates the matching score as a product of the business score and the negotiation feasibility score. The matching score may be calculated by other methods such as adding the business score and the negotiation feasibility score or calculating the average value.

The operation of the business matching system 30 of this embodiment will be described. Figure 16 is a diagram showing the operation flow of the business matching system 30.

The operations from steps S21 to S24 in Fig. 16, in which the acquisition unit 11 acquires the needs text and company information and the business distance calculation unit 23 calculates the importance of each phrase, are the same as the operations from steps S11 to S14 in the first embodiment shown in Fig. 8. Therefore, hereinafter, only the operations from step S25 in Fig. 16 onwards, in which the business score is calculated, will be described.

16, when the importance of each phrase is calculated, the business distance calculation unit 23 calculates the closeness between the needs of the matching target company and the needs of the candidate companies as a business score (step S25). The business distance calculation unit 23 calculates, for example, the Euclidean distance between the needs of the target company and the needs of each of the candidate companies, and calculates the similarity between the needs of the target company and the needs of each of the candidate companies as a business score.

Once the business score has been calculated, the feature calculation unit 41 of the negotiation feasibility score calculation unit 31 converts the company information relating to the target company and the candidate company into features (step S26). The feature calculation unit 41 extracts information on the needs type, desired matching industry, and industry from the needs text of the target company as company information. The feature calculation unit 41 also extracts information on the desired matching industry and industry from the company information of the candidate company as company information. The feature calculation unit 41 uses the extracted company information as input and converts the information of each item into features using a learning model.

Once the information for each item has been converted into features, the negotiation similarity calculation unit 42 calculates a negotiation feasibility score using the similarity calculated based on the features calculated by the feature calculation unit 41 (step S27). The negotiation similarity calculation unit 42 uses, for example, the similarity with past negotiation history as the negotiation feasibility score. The negotiation similarity calculation unit 42 may calculate the negotiation feasibility score by using cosine similarity to calculate the similarity with past negotiation history, and calculate the negotiation feasibility score as the product of the success rate of similar past negotiations and the similarity.

Once the negotiation feasibility score is calculated, the matching unit 32 calculates a matching score based on the business score and the negotiation feasibility score. The matching unit 32 calculates the matching score as the product of the business score and the negotiation feasibility score (step S28).

Once the matching score has been calculated, the output unit 15 outputs the matching result based on the matching score (step S29). The output unit 15 may, for example, output the names and matching scores of a preset number of candidate companies in descending order of matching score. Figure 17 is a schematic diagram showing an example of a display screen on which candidate companies for matching are output in descending order of matching score. In the example of Figure 17, the company names and matching scores of companies with high matching scores are displayed along with the values of the negotiation feasibility score and business score.

The output unit 15 may also output, for example, the name of the candidate company with the highest similarity and the matching score. The output unit 15 may output the name of the candidate company, the matching score, and one or both of the company information and the company information of the candidate company. The output unit 15 may also output the name of the candidate company with a matching score equal to or higher than a preset standard and the matching score. The output unit 15 may also output the name of the candidate company, the matching score, and one or both of the text and company information used in the response by the candidate company. The output unit 15 may also be a display control unit that controls a display device to display the information outputted above.

When outputting the matching results, an operation button for contacting candidate companies for matching may be displayed on the display screen. FIG. 18 shows an example of a display screen displaying an operation button for contacting candidate companies for matching. In FIG. 18, when an operator clicks, for example, the "Contact Company D" button with a mouse, an email can be sent to company D. When communicating with candidate companies by operating the display screen, for example, the business matching system is configured to include a communication unit as shown in FIG. 19. FIG. 19 shows a configuration in which business matching system 50 includes a communication unit 16 in addition to the same configuration as business matching system 30 in FIG. 12.

The business matching system 30 of this embodiment calculates a matching score using the similarity and importance of phrases extracted from the needs text of the matching target and the company information of the candidate company, as well as the success rate based on past business negotiation results. Therefore, the business matching system 30 of this embodiment can more accurately present candidate companies for matching.

Third Embodiment
The third embodiment of the present invention will be described in detail with reference to the drawing. Fig. 20 is a diagram showing an outline of the configuration of a business matching system 100 of this embodiment.

The business matching system 100 of this embodiment is a system that matches businesses between companies. The business matching system 100 of this embodiment includes a matching unit 101. The matching unit 101 outputs a matching score between the first company and the second company based on the similarity between a natural language sentence indicating the needs of the first company and company information including the industry related to the second company and product information related to the products handled by the second company. Specifically, the matching unit 101 calculates a matching score based on the similarity between a natural language sentence indicating the needs of the first company, which is a target company for matching, and company information of the second company, which is a candidate company for matching. In addition, the matching score is an index indicating whether the second company, which is a candidate company for matching, is suitable as a matching destination for the first company, which is a target company for matching.

The matching unit 14 of the first embodiment and the matching unit 32 of the second embodiment are each an example of the matching unit 101 of this embodiment. Furthermore, the matching unit 101 of this embodiment is one aspect of a matching means.

The operation of the business matching system 100 of this embodiment will be described. FIG. 21 is a diagram showing the operation flow of the business matching system 100 of this embodiment. The matching unit 101 outputs a matching score between the first company and the second company based on the similarity between a natural language sentence indicating the needs of the first company and company information including the industry related to the second company and product information related to the products handled by the second company (step S101).

The business matching system 100 of this embodiment outputs a matching score between a first company and a second company based on the similarity between a natural language sentence indicating the needs of the first company and the company information of the second company. Therefore, the business matching system 100 of this embodiment can present suitable matching candidates without requiring past transaction data for matching. Therefore, the business matching system 100 of this embodiment can present suitable business partner candidates regardless of the presence or absence of transaction data.

Each process in the business matching system of the first to third embodiments can be performed by executing a computer program on a computer. FIG. 22 shows an example of the configuration of a computer 200 that executes a computer program that performs each process in the business matching system of the first to third embodiments. The computer 200 includes a CPU (Central Processing Unit) 201, a memory 202, a storage device 203, an input/output I/F (Interface) 204, and a communication I/F 205.

The CPU 201 reads out and executes computer programs for performing each process from the storage device 203. The CPU 201 may be configured by a combination of a CPU and a GPU (Graphics Processing Unit). The memory 202 is configured by a DRAM (Dynamic Random Access Memory) or the like, and temporarily stores the computer programs executed by the CPU 201 and data being processed. The storage device 203 stores the computer programs executed by the CPU 201. The storage device 203 is configured by, for example, a non-volatile semiconductor storage device. Other storage devices such as a hard disk drive may be used for the storage device 203. The input/output I/F 204 is an interface that accepts input from the worker and outputs display data, etc. The communication I/F 205 is an interface that transmits and receives data between each device constituting the business matching system and the user's terminal, etc.

The computer programs used to execute each process can also be stored on a recording medium and distributed. As the recording medium, for example, a magnetic disk such as a magnetic tape for recording data or a hard disk can be used. As the recording medium, an optical disk such as a CD-ROM (Compact Disc Read Only Memory) can also be used. A non-volatile semiconductor memory device can also be used as the recording medium.

Some or all of the above embodiments may also be described as, but are not limited to, the following notes:

[Appendix 1]
A business matching system for matching businesses between companies, comprising:
A business matching system comprising: a matching means for outputting a matching score between a first company and a second company based on the degree of similarity between a natural language sentence indicating the needs of a first company and company information including an industry of the second company and product information regarding products handled by the second company.

[Appendix 2]
2. The business matching system according to claim 1, further comprising: an output unit for outputting matching candidate companies based on the matching score.

[Appendix 3]
The business matching system according to claim 2, further comprising a communication means for communicating with the matching candidate companies.

[Appendix 4]
A natural language sentence including the needs and an acquisition means for acquiring the company information;
and an extraction means for extracting phrases in a predetermined field from the natural language sentence including the needs and the company information,
4. A business matching system according to any one of claims 1 to 3, wherein the matching means calculates the matching score based on the similarity of phrases extracted by the extraction means.

[Appendix 5]
The business matching system of claim 4, wherein the extraction means calculates the probability that a natural language sentence indicating the need and a phrase contained in the company information belong to the specified field, and extracts phrases based on the probability.

[Appendix 6]
further comprising importance calculation means for calculating the importance of the phrase extracted by the extraction means in the predetermined field,
The business matching system according to claim 4 or 5, wherein the matching means outputs the matching score based on the similarity and the importance.

[Appendix 7]
The method further includes a business negotiation feasibility score calculation means for calculating a business negotiation feasibility score based on the type and attributes of the needs of the first company, the type and attributes of the needs of the second company, and past business negotiation results,
A business matching system according to any one of appendices 1 to 6, wherein the matching means further uses the negotiation feasibility score to calculate the matching score.

[Appendix 8]
A business matching system according to any one of appendices 1 to 7, wherein the company information of the second company includes a sentence in a natural language.

[Appendix 9]
A business matching method for matching businesses between companies, comprising the steps of:
A business matching method that outputs a matching score between a first company and a second company based on the similarity between a natural language sentence indicating the needs of the first company and company information including the industry of the second company and product information regarding products handled by the second company.

[Appendix 10]
The business matching method according to claim 9, further comprising outputting candidate matching companies based on the matching score.

[Appendix 11]
The business matching method according to claim 10, wherein communication is performed with the matching candidate companies.

[Appendix 12]
Acquiring a natural language sentence including the needs and the company information;
extracting phrases in a predetermined field from the natural language sentence containing the needs and the company information,
12. The business matching method according to any one of claims 9 to 11, wherein the matching score is calculated based on the similarity of the extracted phrases.

[Appendix 13]
A business matching method as described in Appendix 12, which calculates the probability that a natural language sentence indicating the need and a phrase contained in the company information belong to the specified field, and extracts phrases based on the probability.

[Appendix 14]
Calculating the importance of the extracted phrases in the predetermined field;
The business matching method according to claim 12 or 13, wherein the matching score is output based on the similarity and the importance.

[Appendix 15]
Calculating a possibility of realizing a business negotiation as a business negotiation success score based on the type and attributes of the needs of the first company, the type and attributes of the needs of the second company, and past business negotiation results;
A business matching method described in any one of appendix 9 to 14, further using the negotiation feasibility score to calculate the matching score.

[Appendix 16]
16. The business matching method according to any one of appendices 9 to 15, wherein the company information of the second company includes a sentence in a natural language.

[Appendix 17]
A program recording medium having a business matching program recorded thereon for matching businesses between companies,
A program recording medium having recorded thereon a business matching program that causes a computer to execute a process of outputting a matching score between a first company and a second company based on the similarity between a natural language sentence indicating the needs of a first company and company information including the industry of the second company and product information regarding the products handled by the second company.

[Appendix 18]
A program recording medium recording the business matching program described in Appendix 17, which records a program that causes a computer to execute a process of outputting matching candidate companies based on the matching score.

[Appendix 19]
A program recording medium as described in Appendix 18, which records a business matching program that causes a computer to execute a process of communicating with the matching candidate companies.

[Appendix 20]
A process of acquiring a natural language sentence including the needs and the company information;
A process of extracting phrases in a predetermined field from the natural language sentence containing the needs and the company information, respectively;
20. A program recording medium according to any one of appendices 17 to 19, having recorded thereon a business matching program for causing a computer to execute a process of calculating the matching score based on the similarity of extracted phrases.

[Appendix 21]
A program recording medium as described in Appendix 20, which records a business matching program that causes a computer to execute a process of calculating the probability that a natural language sentence indicating the need and a phrase contained in the company information belong to the specified field, and extracting phrases based on the probability.

[Appendix 22]
A process of calculating the importance of the extracted phrases in the predetermined field;
22. The program recording medium according to claim 20 or 21, having recorded thereon a business matching program that causes a computer to execute a process of outputting the matching score based on the similarity and the importance.

[Appendix 23]
Calculating a possibility of realizing a business negotiation as a business negotiation success score based on the type and attributes of the needs of the first company, the type and attributes of the needs of the second company, and past business negotiation results;
A program recording medium described in any one of Appendices 17 to 22, recording a business matching program that causes a computer to execute a process of calculating the matching score further using the negotiation success/failure score.

[Appendix 24]
24. The program recording medium of any one of appendices 17 to 23, wherein the company information of the second company includes a sentence in a natural language.

The present invention has been described above using the above-mentioned embodiment as an exemplary example. However, the present invention is not limited to the above-mentioned embodiment. In other words, the present invention can be applied in various aspects that can be understood by a person skilled in the art within the scope of the present invention.

REFERENCE SIGNS LIST 10 Business matching system 11 Acquisition unit 12 Extraction unit 13 Importance calculation unit 14 Matching unit 15 Output unit 16 Communication unit 21 Phrase extraction unit 22 Business expression determination unit 23 Business distance calculation unit 30 Business matching system 31 Negotiation feasibility score calculation unit 32 Matching unit 41 Feature amount calculation unit 42 Negotiation similarity calculation unit 43 Business distance calculation unit 44 Matching score calculation unit 50 Business matching system 100 Business matching system 101 Matching unit 200 Computer 201 CPU
202 Memory 203 Storage device 204 Input/output I/F
205 Communication I/F

Claims (10)

A business matching system for matching businesses between companies, comprising:
An acquisition means for acquiring company information including a natural language sentence including the needs of a first company and a type of business of a second company and product information regarding products handled by the second company;
an extraction means for extracting phrases in a predetermined field from the natural language sentence including the needs and the company information;
a matching means for outputting a matching score between the first company and the second company based on a similarity between a phrase extracted from a natural language sentence including the needs and a phrase extracted from the company information ;
A business matching system. The business matching system according to claim 1 , further comprising an output unit for outputting matching candidate companies based on the matching score. The business matching system according to claim 2 , further comprising a communication means for communicating with the matching candidate companies. A business matching system as described in any one of claims 1 to 3 , wherein the extraction means calculates a probability that a natural language sentence indicating the need and a phrase contained in the company information belong to the specified field, and extracts phrases based on the probability. further comprising importance calculation means for calculating the importance of the phrase extracted by the extraction means in the predetermined field,
5. The business matching system according to claim 1, wherein the matching means outputs the matching score based on the similarity and the importance. The method further includes a business negotiation feasibility score calculation means for calculating a business negotiation feasibility score based on the type and attributes of the needs of the first company, the type and attributes of the needs of the second company, and past business negotiation results,
The business matching system according to claim 1 , wherein the matching means calculates the matching score by further using the negotiation feasibility score. A business matching system for matching businesses between companies, comprising:
a matching means for outputting a matching score between the first company and the second company based on a similarity between a natural language sentence indicating the needs of the first company and company information including an industry of the second company and product information regarding products handled by the second company;
a business negotiation feasibility score calculation means for calculating a business negotiation feasibility score based on the type and attributes of the needs of the first company, the type and attributes of the needs of the second company, and past business negotiation results;
Equipped with
The matching means further uses the negotiation possibility score to calculate the matching score.
Business matching system. The business matching system according to any one of claims 1 to 7, wherein the company information of the second company includes text in a natural language. A business matching method for matching businesses between companies, comprising the steps of:
The computer
Acquire natural language text including the needs of a first company and company information including an industry of a second company and product information regarding products handled by the second company;
extracting phrases in a predetermined field from the natural language sentence containing the needs and the company information,
A business matching method that outputs a matching score between the first company and the second company based on the similarity between a phrase extracted from a natural language sentence containing the needs and a phrase extracted from the company information . A business matching program that matches businesses between companies,
A process of acquiring natural language text including needs of a first company and company information including an industry of a second company and product information regarding products handled by the second company;
A process of extracting phrases in a predetermined field from the natural language sentence containing the needs and the company information, respectively;
A business matching program that causes a computer to execute a process of outputting a matching score between the first company and the second company based on the similarity between a phrase extracted from a natural language sentence containing the needs and a phrase extracted from the company information .

Images