Commercial Transaction Agent System

Abstract

To provide an agent system operation method that effectively supports a matching evaluation process by inputting attributes of a selling company and a purchasing company, attributes of a product, and importance parameters in the decision-making of the companies into an agent system, in relation to B-to-B (business-to-business) e-commerce technology.SOLUTION: A method of operating an agent system 4, which is a method of performing matching evaluation by quantifying attributes of a selling company and a purchasing company and attributes of a trading product, includes: inputting an importance parameter in decision-making of one company or both companies into the agent system; after mathematically converting all numerical values of the attributes of the selling company and the purchasing company and the attributes of the trading product according to the degree of importance, by using the converted numerical values to perform a multidimensional and simplified matching evaluation process; and outputting company attributes or product attributes that are individual factors that hinder matching.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to electronic commerce technology, and more particularly to technology applied to an agent system that effectively supports BtoB (business-to-business) commerce. [Background technology]

[0002] The agent system here refers to a system that acts between sellers and buyers to make platform businesses, which have been studied and defined in non-patent document 1 and elsewhere, more productive and efficient, as "businesses that stimulate transactions between third parties and provide a foundation for starting new businesses through the supply of products and services that anyone can receive under clear conditions."

[0003] There are various methods for matching to connect sellers and buyers in commercial transactions, such as bulletin board methods, management staff methods, automatic matching methods, etc. In particular, in commercial transactions between companies and individuals (so-called BtoC marketing), bulletin board methods are mostly used, and AMAZON is a famous example.

[0004] In BtoC, from the seller's perspective, there are a large number of target buyers, but since the buyers are individuals and they pay for the products (mainly consumer goods), it is said that marketing activities can be focused on specific targets. For this reason, marketing such as creating web pages that emphasize taste and feeling allows commercial transactions to be concluded with quick decision-making (so-called one-click).

[0005] On the other hand, in BtoB marketing, the products (industrial goods and capital goods) and services are often very expensive compared to BtoC, and therefore the decision to purchase the product is not made by an individual, but by many people belonging to an organization (mainly within a company). Furthermore, at various stages of the process within a company related to the purchase, for example, the department that decides the required specifications of the product, the department that uses the product, and the department that purchases the product, each have different requirements for the product, and the scope in which requirements can be determined as a department is also limited. Furthermore, each department usually has a person in charge and multiple levels of decision-makers. For this reason, even if one person in a company views a web page for electronic transactions prepared by a sales company, it generally does not lead to a commercial transaction. In other words, it is clear that a business that is only intended to display a web page for electronic transactions prepared by a sales company, as in BtoC marketing, is difficult to establish in BtoB marketing.

[0006] Taking Non-Patent Document 2 as an example, through a website called TTPP by JETRO, case registration and case introductions are simultaneously sent in both Japanese and English, and it seems that business opportunities can be expanded globally. However, on such business matching sites, even if you are interested in a case, you cannot check detailed information on the site. Therefore, people from each company must proceed with detailed consideration by exchanging information by e-mail. Doing business in this way while targeting many companies is a waste of time and labor, and it is extremely difficult to do large-scale transactions with overseas companies in particular. Therefore, transactions through such websites are limited to very small-scale transactions, such as the sale of small standard parts or used machine tools.

[0007] In other words, as mentioned above, when trying to conduct large or medium-sized B2B electronic commerce using bulletin boards, it has not been possible to simplify and speed up business information exchange and decision-making.

[0008] In addition, in B2B e-commerce using bulletin boards, the lengthy internal decision-making procedures and the wide range of people involved in a company, as well as the difficulty for foreign companies in understanding the high-context communication of Japanese companies, are potential obstacles. For example, companies that are familiar with domestic Japanese trade practices often conduct business transactions with minimal clear indication of purchase requirements based on a mutual tacit understanding based on past business results, even if not all purchase requirements are clearly stated in detail. However, it is difficult for low-context foreign companies, which prefer direct and easy-to-understand expressions, to conduct e-commerce transactions with Japanese companies with minimal clear indication of purchase requirements. Naturally, it is necessary for global business transactions to clearly state basic contracts, detailed contracts, applicable laws, regulations, and standards, and detailed product specifications, and to make them mutually visible, so it is clear that a business that is only conducted by disclosing web pages will not easily be established in global B2B marketing.

[0009] Next, we will describe commercial transactions using the management staff method and the automatic matching method. The management staff method involves holding events such as trade fairs to bring together buyers and sellers and provide a forum for interviews. Commercial transactions here are mainly conducted by the participants of the event alone, and as direct involvement and promotion by the management staff cannot be expected, the outcome is left to chance. For this reason, there is little significance in using a commercial transaction agent system aimed at electronic transactions.

[0010] Although it is not a commercial transaction between companies, there is already a matching technology known as a dating support system for automatic matching. Patent Document 4 describes an example of a technology in which an operator intervenes between two people who wish to meet and provides automatic matching. In this technology, "the dating support information processing device comprises a user information storage means for storing user information on a plurality of users, a user selection means for selecting other users who are likely to like each user based on the user information stored in the user information storage means, and a transmission means for transmitting information on the other users selected by the user selection means to an information terminal device of the user." Using this technology, the operating site proposes matching.

[0011] Furthermore, as in Patent Document 3, a technology is proposed to improve the matching rate by adding not only user information about the user but also behavioral history information of the user collected by the operating site, as follows: "The marriage support method includes a first step in which a processing means extracts from the behavioral history database an opposite-sex person who has a behavioral history of making a dating request to the target person as a first opposite-sex linker, a second step in which a same-sex person who has a behavioral history of making a dating request to the extracted first opposite-sex linker is extracted from the behavioral history database as a first same-sex linker, and a third step in which an opposite-sex person who has a behavioral history of making a dating request to the extracted first same-sex linker is extracted from the behavioral history database as a recommended candidate."

[0012] However, these examples are matching technologies based on the applicant's past and current attributes, and are nothing more than a technology that performs cluster analysis to gather similar people together and determines which clusters are likely to combine. In other words, the operating sites do not consider suggestions for improvements such as how to make a match between clusters that are not easily combined. For this reason, with the technology presented here, many possible matches are determined by the program to be "unusable" and discarded. For these reasons, well-known online dating matching technologies cannot be applied to B2B commercial transactions.

[0013] In addition, in Patent Document 2, as a prototype ordering and receiving intermediation system, one purchaser requests multiple potential recipients to send information on prototypes that can be accepted and information on the number of available jobs, and the evaluation score of each potential recipient is calculated based on the multiple potential recipients' information and item coefficients. A list of potential recipients is created in which the multiple potential recipients are arranged in descending order of evaluation score, and the list is displayed on the screen of the purchaser's terminal. "A server is interposed between the purchaser and multiple potential recipients, and the server weights and sums the prices and delivery dates of each potential recipient and compares the magnitude of the numerical values ​​to rank the potential recipients. As a result, the purchaser can easily select the most suitable contractor for the prototype production he or she wishes to request, taking into account the balance between price and delivery date." Also, "The contractor information items are not limited to contractor field information and congestion information, and can be freely set. For example, contractor evaluation information, which is the aggregated result of evaluations given by past purchasers, can be included as a contractor information item." In order to realize "the trust in quality and the desire to do business with a trustworthy contractor as a trading partner," for example, "when or before the prototype production transaction is completed, the purchaser can select one of the evaluation stages, such as 'Very Satisfied,' 'Satisfied,' 'Average,' or 'Unsatisfied,' and enter the evaluation stage into the contractor information." This could already make the person who entered the information the gatekeeper for deciding on the purchaser. In other words, there is a high possibility that the person who entered the information could even destroy the possibility of business transactions with some new and excellent candidates for the contract. This has the major disadvantage of not promoting competition and innovation within the horizontal division of labor using the business platform, and depriving participants of the opportunity for co-creation.

[0014] In this way, the score calculation means calculates the score of each candidate for an order request based on the contractor information and order request information, by multiplying the numerical values ​​corresponding to items such as price, delivery date, and orderer evaluation by their respective weighting coefficients and using a weighted sum polynomial. This is not an evaluation of matching, but simply an evaluation of economic consideration, and is a field different in purpose and function from the present invention.

[0015] Patent Document 5 also proposes a method of automatically selecting quotation information stored in an quotation information storage means in the order of priority of price and delivery date in order to adjust the balance between price and delivery date. As an algorithm for determining the lending request destination, P is the price, D is the number of days past the delivery date, and w is a weighting percentage that is appropriately set according to how much importance / consideration is attached to the delivery date delay, and the total of the price and the opportunity loss amount due to the delivery date delay is evaluated as P+(D×P×w), which is reasonable. However, this technology, like Patent Document 2, quantifies the comparison items to be equivalent to the amount, weights them, and adds them up, and the one with the smaller total is favored, which is simply an evaluation of economic efficiency, and is in a field different in purpose and function from the present invention.

[0016] The example in Patent Document 1 does not provide automatic matching for BtoB, but provides "agent system technology that autonomously performs commercial transactions" in BtoC. Here, a life agent is an entity that autonomously obtains various information useful to the consumer and records various activity histories. An autonomous life agent is placed for the consumer as a buyer, and an autonomous business agent (with the function of conducting business) is placed for the company as a seller, and they are connected to each other via a network. By having a content aggregator on the consumer side, the transaction form is completely different from the conventional bulletin board method, and the life agent narrows down the vague requests of the buyer to appropriate products, communicates with the business agent, and starts business with the seller. In this way, it is said that the technology realizes a BtoC environment that truly contributes to consumers by autonomously and automatically conversing and collaborating. However, since the transaction here is basically one-to-one, automatic matching technology for one-to-many transactions is not disclosed. The inter-agent protocol and audit agent are for monitoring the transaction, and the life agent does not evaluate the seller company. Therefore, it must be said that applying this automatic matching technology to BtoB would be difficult. [Prior art documents] [Patent documents]

[0017] [Patent Document 1] Patent application 2017-517397 [Patent Document 2] Patent Publication No. 2007-219879 [Patent Document 3] Patent Publication No. 2017-045342 [Patent Document 4] Patent Publication No. 2002-207842 [Patent Document 5] Patent Publication No. 2002-169991 [Non-patent literature]

[0018] [Non-Patent Document 1] Hiroshi Mizoshita, "Research Review on Platform Business and Business Models," Hiroshima University Management Review, Vol. 18, 2017 [Non-Patent Document 2] "TTPP: An inquiry database that connects you with overseas companies" issued by the Japan External Trade Organization (JETRO) TTPP Secretariat Summary of the Invention [Problem to be solved by the invention]

[0019] Until now, global B2B e-commerce has mainly been conducted via internet screens, where sellers and buyers view each other's public information, and matching has been done with the uncertainty of chance encounters and with a high level of information constraints, with an emphasis on price and delivery time. Going a step further, technology has been introduced in which a system obtains information on the prices and delivery times of sellers' products, and uses that data to automatically perform economic calculations and rank sellers. However, this technology also resulted in a simple and superficial matching evaluation, and was unable to create the mutual trust and long-term stable business environment that has been cultivated in traditional commercial transactions, where each company makes the best proposal and the best products emerge through mutual effort.

[0020] In other fields, various methods have been proposed as matchmaking technologies for dating, but they are matching technologies based on past and current attributes, and are ultimately nothing more than a technology that performs cluster analysis to gather similar things together and finds which clusters are likely to combine. Therefore, they cannot be applied to B2B transactions that create a place where heterogeneous things can cooperate with each other as a horizontal division of labor.

[0021] Previous technology was not able to provide an agent system that could perform BtoB matching rationally and efficiently, and create a business environment in which selling and purchasing companies could each make their best proposals and create the best products through mutual efforts, without being restricted by general matching evaluation methods. [Means for solving the problem]

[0022] One of the reasons why each company has its own unique policy and decision-making process is that it assigns its own priorities (importance levels) to various business issues. With this in mind, the commerce agent system adjusts the attributes (numerical values) of both companies and the attributes (numerical values) of the trading products with the importance levels (numerical values) of each other to evaluate the match. Furthermore, by clarifying which attributes match and which do not, it actively works to improve the match.

[0023] This method is characterized by a method for evaluating matching by quantifying the attributes of the selling and purchasing companies and the attributes of the traded goods for the purpose of BtoB (business-to-business) electronic commerce, in which importance parameters in the decision-making of one or both companies are input into an agent system, the numerical values ​​of the attributes of the selling and purchasing companies and the attributes of the traded goods are all mathematically converted using the importance, and then a multi-dimensional and simplified matching evaluation process is performed using the converted numerical values, and further the company attributes or product attributes that are individual factors impeding matching are output. Effect of the Invention

[0024] According to the present invention, in regard to the technology of BtoB (business-to-business) electronic commerce, by inputting the attributes of the selling company and the purchasing company, the attributes of the product, and the importance parameters in the decision-making of the companies into the agent system, it is possible to effectively and accurately provide a matching evaluation process. Furthermore, the present invention can provide an agent system that forms a business field where each company can make the best proposal and create the best product through mutual efforts by outputting the matching evaluation process as a tailor-made specific solution for each company without being restricted by the matching evaluation process as a general solution. Furthermore, improvements such as information correction and addition to realize stronger matching between user companies can be made by the operator through an information terminal, or through the agent, enabling the creation of a global user company. [Brief description of the drawings]

[0025] [Figure 1] 1 is a diagram showing an overview of a configuration example of an agent system according to an embodiment of the present invention; [Diagram 2] FIG. 13 is a diagram showing an example of global business hierarchical data. [Diagram 3] This is a graph adjusted based on basic information and importance of the selling companies. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] FIG. 1 shows an example of the system configuration. The agent system 4 is composed of a storage device, a server, an input / output device, and the like, and is connected by a system network 13. The company introduction information of each company 1 in the domestic group 2 and each company 1 in the overseas group 3, including business terminology, technical terminology, legal terminology, etc., is managed in UTX format, etc., and further machine-translated or manually translated into the language of each country, and stored as company data 5 as the company attributes of each company. In addition, business practices and business-related laws and regulations, environmental regulations, industrial standards, etc., which differ from country to country, are linked to each target commercial transaction and stored as knowledge data 7. Such basic data is protected by IDs and passwords, processed by the agent server 8, and can be viewed by the companies 1 via the network 12. The results of the matching diagnosis output by the matching evaluation process 9 are output to the company information input / output device 11, and business negotiations can be conducted based on that information with each organization or executive of each company by methods 14 such as telephone, email, and interviews.

[0027] In a vertically integrated business system, a general trading company or a top company that performs final assembly has managed the entire process from upstream to downstream and integrated the supply chain. Many of the people involved in the company that plays the role of integrating the supply chain have traveled overseas to investigate local conditions, held long meetings with the people involved, and concluded contracts for business transactions. This is how legacy businesses (traditional business processes) were established. However, with the increase in digitalization and horizontal division of labor businesses like today, companies can no longer expect to grow with traditional legacy businesses alone. E-commerce has become an indispensable means for coexistence with each company in the horizontal division of labor business arena and the creation of groundbreaking products, but it is a well-known fact that simply digitizing (digital transformation) the processes of legacy businesses does not solve the problem. However, as long as businesses are run by people, even important decision-making processes will not be mechanized, and in high-value business transactions, it is thought that legacy processes by many people involved in the company (executives, design, manufacturing, procurement, production management, sales, etc.) will continue to play a key role in decision-making.

[0028] Generally, companies attach great importance to efforts toward the SDGs in order to solve social issues together. Therefore, in high-value commercial transactions, it goes without saying that companies make decisions on commercial transactions after confirming the track record of efforts toward the SDGs, along with the company policies, business performance, technical capabilities, quality maintenance capabilities, and price competitiveness of potential business partners. However, in actual commercial transactions, the degree of interest in the SDGs is not uniform from company to company. In other words, the conditions and levels of business matching cannot be applied uniformly to all companies. It is necessary to measure the quantitative importance of items related to decision-making for each company, and even for each organization within the company, and reflect them in the conditions of business matching. Here, the importance information specific to each company is surveyed and set, and is stored in the company data5 as the importance attribute data for each company.

[0029] Data on past and currently ongoing commercial transactions (such as the start date of the transaction, product name, specifications, price, applicable laws and regulations, applicable industrial standards, defect rate, number of delivery delays, contribution to cost reduction, logistics company, etc.) is stored in transaction data 6. This data can also be used to select details of the buyer's purchasing requirements and the seller's sales requirements, and used in matching evaluation processing for new commercial transactions. For example, even if the buyer's presentation of purchasing requirements is insufficient for the seller's perspective, it can be supplemented to some extent with information from this transaction data.

[0030] In addition, the agent, who acts as a strategic partner between the buyer and the seller, collects the aforementioned corporate attribute data through individual interviews with the target company's contact organization, executives, design, manufacturing, procurement, production management, sales, etc., in order to influence the legacy process within the target company. By understanding the context during conversations with company personnel and inputting the results of that information accurately via the company information input / output device 11, the automatic matching evaluation processing technology using AI technology can be effectively used. Furthermore, the final goal is not to find potential trading partners using the matching evaluation processing technology. Using the matching diagnosis output by the matching evaluation processing 9, the agent clearly explains the issues of concern to both companies and makes suggestions to improve the matching between the buyer and seller companies.

[0031] Furthermore, such proposals do not function simply by transmitting inorganic, electronic information. Based on the output information, the agent makes strategic proposals to decision makers in both the seller and buyer companies, encourages effective improvements, and if they are agreed to, the information is considered to have been transmitted to the target companies. To this end, the agent system quantifies the items in which the matching is inconsistent through a matching evaluation process 9, and performs a process of notifying the agent in detail of this information through the company information input / output device 11. In this way, the agent and the agent system take measures to address the matching inconsistency, and after taking measures, go through the process of updating each company's attributes and product attributes, providing BtoB commerce with the possibility that each company 1 can make the best proposal and create the best product through mutual efforts.

[0032] Here, we will explain the preparations for the matching evaluation process. Based on past commercial transaction results over a certain period of time, the information of the product required by the buyer company is automatically compared with the information of the product provided by the seller company, and the attributes of the buyer company are also automatically compared. The results of these comparisons are classified from the higher concepts of the company's commercial transaction requirements to the lower concepts, such as strategic, administrative, operational, and prerequisites, and are temporarily stored as global business hierarchical data 10 classified as shown in Figure 2. This hierarchical data can be further assigned medium and small classifications. Here, the medium classifications for SDGs, communication, and quality standards are displayed as examples. By creating a hierarchy, if an obstacle or issue in matching is discovered, it is easy to infer which internal organization should be approached, which is extremely important in advancing problem solving.

[0033] Next, as a continuation of the preparation for the matching evaluation process, an example of a method for constructing parameters for the matching judgment process will be described. All items of the product attributes and company attributes of the buyer and seller obtained in the course of commercial transactions over a certain period of time in the past are quantified and then subjected to multivariate analysis. When corporate data is stratified as shown in Figure 2, it is thought that data in higher-level concepts is mostly related to corporate attributes, while data in lower-level concepts is mostly related to product attributes.As an example of quantification, each basic information item of a company's attributes is entered as an integer from 1 to 10. The most complete attribute value observed from the other company is set to 10, and the most incomplete attribute value observed from the other company is set to 1. As a guideline, a value of 7 or more is a good value, and a value of 5 or less is an evaluation below average. Next, the cases in which the business transaction was concluded are classified as (1) cases in which it went smoothly, (2) cases in which the terms were changed or negotiations were made, and (3) cases in which it was ultimately unsuccessful even after negotiations, and a discriminant analysis is performed on the quantified data. Furthermore, the factors are narrowed down by cluster analysis. Next, from the perspective of the buyer company, a multiple regression analysis is performed on those factors with the required value satisfaction of the seller's product as the objective variable, and a multiple regression analysis is performed with the reliability of the seller's company as the objective variable. In order to avoid multicollinearity, it is desirable to exclude one of the basic information items that are considered to be highly correlated with each other as explanatory variables.

[0034] By setting four threshold levels for each axis, it is possible to classify negotiation results and negotiation conditions according to combinations of satisfaction and reliability in a decision table, as shown in the table below. This table was created from the perspective of the buyer company, but in a similar way, a decision table can also be created from the perspective of the seller company. However, in general markets, buyers have the initiative in transactions, so there are few opportunities to use this. [Table 1]

[0035] Also, instead of multiple regression analysis, it is possible to quantify the items of product information and company attributes of buyers and sellers, and use each variable as a membership function to calculate the value of the antecedent, and then use fuzzy inference to derive the product's satisfaction level and company reliability evaluation value as the consequent. After defuzzification, it is possible to create a similar decision table. Furthermore, it is also possible to create a similar decision table by quantifying the items of product information and company attributes of buyers and sellers, and using tensorflow to recognize patterns in all the data, and then grouping them into 16 patterns.

[0036] With the above preparations, a matching evaluation can be performed from the standpoint of an average buyer company using a decision table based on the company data 5 and the transaction data 6. However, since each company has a different mission, management philosophy, and code of conduct, naturally each company places a different degree of importance on each requirement in a business transaction. For this reason, more accurate results can be obtained by performing a matching evaluation based on the unique company attributes corrected for each requirement of each company by their respective importance. Here, as an example, an importance value ranging from minus 5 to plus 5 is input as a parameter for the importance of each item of basic information of the buyer company and the seller company. Here, the greater the positive value, the greater the degree of importance, and conversely, a negative value indicates the degree of less importance. An importance value of 0 is the average level, and the default value can be set to minus 2 or minus 3. It is desirable to revise this decision table at an appropriate frequency. Changes in the business environment also affect company attributes and importance, and as a result, the contribution rate of each attribute value in the matching evaluation also affects it. EXAMPLES

[0037] Table 2 shows the results of a survey of the executives, design department, manufacturing department, and procurement department of a certain company on the importance they place on the attribute values ​​of their business partners. The importance varied depending on the job, but the maximum value selected for each was taken to represent the consensus on the importance of the entire company. The actual basic information is further broken down, but here we have simplified the explanation by only showing representative items. [Table 2]

[0038] We will explain the conversion method to obtain corporate attribute values ​​corrected for the importance of each requirement of each company (attribute values ​​of trading partners). Matching evaluation processing is easier if the converted value is set to a maximum of 10. If the importance value (value of importance) is greater than the value of the company's attribute value, the converted value will be negative, resulting in an evaluation that is even worse than the most incomplete evaluation.

[0039] As a concrete example, attribute values ​​were converted using the following formula: Here, Company A is the seller and Company B is the buyer. (Formula 1) Value after correction of Company A's importance = (Company A's basic information value - Company B's importance value) x 10 / (10 - Company B's importance value)

[0040] The graph in Figure 3 compares the basic information values ​​of the company attributes of Company A before and after adjustment based on the importance. The importance values ​​here are those obtained in Table 2. Basic information of the 10 items of seller companies (The corporate attributes here are defined broadly to include both the attributes of the trading company and the attributes of the trading product.) The average evaluation score for sellers was 7.9 points before adjustment, but when values ​​were substituted into Equation 1 and adjustments were made based on the importance attributes of buyer companies, the average evaluation score for sellers after adjustment for importance fell to 7.1 points, as shown in the example in Figure 3.

[0041] Clarifying the importance of corporate attributes in this way meant that a large portion of the high-context information within Japanese companies was converted into low-context information. The importance of corporate attributes can be calculated automatically if there are a large number of past transaction results to make them statistically significant. However, when there is a small amount of transaction history, it was possible to quantify the importance by having each organization within the target company answer a simple questionnaire. Furthermore, the importance attribute data for companies and internal organizations was also stored as part of corporate data5.

[0042] Next, using the values ​​of each factor corrected for importance, a multiple regression analysis was performed from the buyer's perspective, with the satisfaction of the seller's product requirements as the objective variable, and a multiple regression analysis with the seller's trust as the objective variable. The converted values ​​were applied to the multiple regression equation mentioned above, and the magnitude relationship between the two obtained values ​​and the set threshold value was used to obtain the result "trust is slightly dissatisfied, requirements are mostly satisfied" as the first matching evaluation using the decision table mentioned above. This result was almost consistent with the results of interviews at the company and the company's internal trading sense. In this way, using the importance of the companies makes it possible to more accurately evaluate the matching between the target companies.

[0043] Based on the converted company attributes and converted product attributes, it was possible to carry out a matching evaluation appropriate for the unique combination between companies. The agent uploaded the results of the first matching evaluation obtained in this way to the seller company, made the information public, and explained the need for improvements to the proposal content, etc., in accordance with the decision table. In particular, the correction values ​​after converting the seller's basic information according to the buyer's importance revealed specific issues.

[0044] If we were to stop the process here, then according to market principles, if the maker's matching rating is low, the price of the product would go down, and if the buyer's matching rating is low, the price of the product would go up. This would not create a win-win relationship, and is not the true intention of the agent system.

[0045] In the matching evaluation process, it was possible to prioritize the items that had a negative impact, which enabled the seller company to focus on the proposed items that needed to be corrected and take swift action. In this example, as is clear from the graph in Figure 3, in the communication item, the seller company's basic information had an attribute value of 6 points, but the buyer company's emphasis value was a high 5 points, so after adjusting for emphasis, the converted value for communication was only 2 points. Since communication was deemed below average, it was determined that the match would not be successful unless the seller company improved its communication performance.

[0046] In the basic information, the attribute value for communication was 6 points, the details of which are shown in (Table 3). The shaded areas in the table are the initial individual attribute values. The attribute value as the average of the seven items was 6.3 points, but "email response period" was 10 days, giving it 4 points, and "minutes" was inaccurate, giving it 3 points, both of which were significantly below the average. [Table 3]

[0047] The important next step was to complete the improvements within the seller company, revise the seller company's corporate attributes or product attributes as a result, upload them to the company data, and then perform the matching evaluation process again. In this example, if the communication aspect was improved to "reply to email inquiries within two days" and the minutes were improved to the level of "ensuring accuracy," the attribute value would rise to 7.6 points. If this was corrected with an emphasis value of 5, the corrected attribute value would be 6.0 points, which is just above the average, resulting in a result of "trust is almost unsatisfactory, requirements are almost satisfied," and the deal was closed. EXAMPLES

[0048] The importance of the buyer company has a large impact on the matching evaluation, but this is not necessarily true in all cases. Also, the importance values ​​for most company attributes are set independently, without influence from other companies. However, there are exceptions, and price (sales amount) is one example. With regard to price, the contribution rates of the buyer company and the seller company to decision-making are usually roughly equal. However, if the seller company has a large market share or is overwhelmingly large in size, the decision-making in determining price may be heavily dependent on the seller company. Therefore, with regard to price, the importance values ​​of the buyer company and the seller company will influence each other.

[0049] We will use the matching evaluation of electrocoating of parts as an example. Prices are displayed in basic units, and are the unit price (Japanese yen) of electrocoating per square meter of parts. We assume that the average unit price is 100 yen, including manufacturing costs, management costs, and minimum profits. According to company data 5 and transaction data 6, in a recent transaction, the seller company offered 110 yen, with an attribute score of 6 and an importance score of 0, while the buyer company offered 110 yen, with an attribute score of 6 and an importance score of 0. Details are shown in (Table 4). [Table 4]

[0050] If the situation had continued, the deal would have been concluded. However, due to a change in policy by the seller company, it became necessary to achieve a profit margin of 20%. As a result, it was discovered that the emphasis on price had changed from 0 points to 2 points. Therefore, the buyer company's attribute value regarding price was corrected and converted using the seller's emphasis. Attribute values ​​can be converted using (Formula 1). In this case, Company A is the buyer company, and Company B is the seller company. As a result of the conversion, the buyer company's attribute value after the correction of the emphasis on price dropped from 6 points to 5 points. 6 points is the lower limit in the matching conditions. In order for the buyer company to obtain an attribute value of 6 points or more after the correction of the emphasis on price, it is necessary to secure 7 points as a base value and obtain 6.3 points after the correction. This is a commercial transaction requirement. As for the purchase price, as shown in [Table 4], the price was revised from 110 yen to 115 yen. By setting the purchase price at 115 yen, the transaction was ultimately concluded. The condition was that the buyer's emphasis on price was changed to minus 2 points. As a result, the seller's revised attribute value can be calculated as a compromise of 6 points. Regarding the reduction in importance to that extent, this result was almost consistent with the results of interviews with the company and the company's internal trading sense.

[0051] In this way, it is important to communicate to the buyer that the importance attribute of the seller company has a large influence on the matching conditions for the issues that have become clear in the matching evaluation process. In the above example of electrodeposition painting, the matching was provisionally made possible by changing the seller's importance value for the price from minus 2 points to minus 3 points, but the business transaction should have been concluded without lowering the importance value by improving the cost of the electrodeposition painting process. In that sense, the significance of the technology of the present invention, which clarifies business issues by clearly indicating the change in the importance value, is great.

[0052] With this technology, the agent was able to provide both the buyer and seller with the results of an intermediate matching evaluation, while at the same time quickly making requests such as relaxing some of the product information for both companies. As a result, quick matching became possible without any difficulty for both the buyer and seller. Furthermore, in the process, it also had the secondary effect of making clear to overseas companies the requirements and priorities that are implicitly understood within Japan. [Industrial Applicability]

[0053] In this way, the technology of the present invention can promote emergent improvements between companies in B2B electronic commerce by rotating the PDSA (Plan-Do-Study-Act) improvement circle in the matching evaluation process. By having an agent as a strategic partner quantify the importance of decision-making and inputting this into an agent system that exists between buyers and sellers, it becomes possible to form a business forum for agreement between companies and emergence through mutual efforts, which can make a significant contribution to the industrial world. [Explanation of symbols]

[0054] 1 company 2. Domestic Group 3. Overseas Group 4. Agent System 5. Company Data 6. Transaction Data 7 Knowledge Data 8 Agent Server 9 Matching evaluation process 10 Global Business Hierarchy Data 11 Corporate information input / output device 12 Network 13 System Internal Network 14. Communication methods such as telephone, email, and interviews

Claims

1. A method for performing a matching evaluation for BtoB (business-to-business) electronic commerce using attribute values ​​that are numerical representations of the attributes of a selling company and a purchasing company and the attributes of a transaction product, and importance values ​​that are numerical representations of the importance that a partner company places on those attributes, the method comprising the steps of: an intermediary (agent) surveys and sets the importance of companies, inputting them as importance values ​​into an agent system; the agent system converts the attribute values ​​of the selling company and the attribute values ​​of the transaction product using a formula with the importance of the purchasing company as a parameter; using the converted attribute values ​​as explanatory variables to perform a multiple regression analysis with product satisfaction and company reliability as objective variables; and performing a matching evaluation process based on the two obtained numerical values ​​and a decision table with product satisfaction and company reliability determined by multiple regression with attribute values ​​obtained from the course of commercial transactions over a certain period of time as explanatory variables as conditions.

2. A method for performing a matching evaluation using attribute values, which are numerical values ​​of the attributes of a selling company and a purchasing company and the attributes of a transaction product, and importance values, which are numerical values ​​of the importance of the other company to those attributes, for the purpose of BtoB (business-to-business) electronic commerce, characterized in that an intermediary (agent) investigates and sets the importance of companies and inputs them as importance values ​​into an agent system, the agent system converts the attribute values ​​of the selling company and the attribute values ​​of the transaction product using a formula with the importance of the purchasing company as a parameter, performs a multiple regression analysis using the converted attribute values ​​as explanatory variables and with product satisfaction and company reliability as objective variables, performs a matching evaluation process based on the two obtained numerical values ​​and a decision table set as conditions for the product satisfaction and company reliability obtained by multiple regression with attribute values ​​obtained from the process of commercial transactions over a certain period of time as explanatory variables, and outputs the importance values ​​of the company attributes or product attributes that are factors that hinder matching.

Images