JP2023147907A - Recommendation device and recommendation method - Google Patents

Abstract

To provide a recommendation device and a recommendation method which recommend an enterprise which becomes a customer candidate.SOLUTION: An open data collection unit 17 collects published or sold business information related to a customer candidate enterprise as open data, and registers the data in a corporation master 13. A business rule conversion unit 16 collects compliance policy of a business entity enterprise, ESG policy, policy of sales and purchase, policy, regulations, etc., and formulates collected each information to register the information in a business rule master 12. An event evaluation unit 15 refers to the business rule master 12 based on past transaction result information and external evaluation data collected by an event information collection unit 14 and registers an evaluation result for each evaluation axis with respect to the customer enterprise in an evaluation master 11. A corporation recommendation unit 18, when there is a recommendation request about the customer enterprise, refers to the corporation master 13, the evaluation master 11 and the business rule master 12 to present an enterprise which can be recommended as a customer.SELECTED DRAWING: Figure 1

Description

The present invention relates to a recommendation device and a recommendation method that follow business rules that differ from company to company and recommend optimal business partners under the rules.

There are various proposals regarding devices and methods for making recommendations regarding products or services and their providers, but for individuals and small and medium-sized enterprises that are close to individuals, there are methods that estimate the preferences of individuals receiving recommendations using various methods. A common method is to recommend products and services that match the individual's tastes (see Patent Document 1). On the other hand, for large companies that engage in a wide variety of transactions, there are proposals to present candidates for product or service providers (clients) (see Patent Document 2), but these proposals fall under the category of "recommendations tailored to individual tastes." Regarding this part, there is no established method yet.

Japanese Patent Application Publication No. 2012-084189 Japanese Patent Application Publication No. 2002-74103

There are various proposals for recommendation devices and methods regarding products or services and their providers (business partners), but as the amount of information that can be collected on the Internet increases, it is important to make recommendations based on the amount of information rather than the amount of information. , suitability and degree of matching to the searcher's (those who receive recommendations) preferences are becoming important. When the searcher is an individual or a small business close to the individual, a method is generally used that uses collaborative filtering technology to determine the recommendation ranking based on the evaluations of searchers with similar preferences.

On the other hand, in large companies (large companies with many business partners, types of transactions, or scale; hereinafter the same applies), these preferences become business rules based on the policies set by the company itself. In large companies, business rules such as regulations are comprehensively formulated to eliminate individual judgment and allow each department within the company to judge the propriety of transactions and the priority of business partners based on unified standards. It is. The present invention provides a device and method for making appropriate recommendations so that a large company can select business partners and conduct transactions based on business rules that exclude personal preferences instead of personal preferences.

In addition to the provisions of laws and regulations, today's large companies also incorporate their own compliance policies, ESG (Environment, Social, Governance) policies, sales and purchasing policies, strategies, policies, regulations, etc. into "business rules" (laws, policies, regulations, etc.). , a generic term that includes all strategies, policies, regulations, etc. (hereinafter used with the same meaning). On the other hand, in large companies, sales and purchasing activities are performed on various information systems, so these business rules essentially need to be incorporated into those information systems and automatically controlled.

However, at present, business rules are not sufficiently incorporated into information systems, and rules are often interpreted and operated manually.

In addition, when taking actions that violate business rules, companies are accountable to their stakeholders for the history and reasons for acting outside of business rules, but currently, the history of actions that violate business rules and the reason will not remain on the system.

Therefore, it is desired to provide a device and a method for recommending business partners with which transactions are permitted and preferred business partners among them, on an information system for conducting the transaction, based on business rules that vary from company to company. In large companies, the business rules themselves are diverse, so it is difficult to imagine that the person in charge of conducting a transaction fully understands all the rules, and without some kind of support system, appropriate This is because it is difficult to select business partners.

When recommending business partners that match the company's business rules, we take into account capital, number of employees, official qualifications and permits held, ESG response status, financial stability, past transaction history with the recommended company, etc. It is necessary to collect various and up-to-date evaluation factors, follow all the latest business rules, and clarify the reasons for the recommendation, then create a list of potential business partners and recommend more desirable business partners. In addition, if a business person attempts to select a business partner that is not recommended, we will encourage them to comply with business rules by reminding them of the precautions they should take when conducting such transactions, and we will also explain the reason for their behavior outside of business rules. It is desirable to require the description or input of

In view of the above circumstances, the present invention formulates business rules that differ for each individual company into a form that can be calculated numerically, collects and digitizes various evaluation information necessary for judgment and evaluation based on the business rules, and provides the following benefits for the company: Recommends the most suitable business partner at that time by automatic calculation (sometimes it shows the judgment result for a single business partner, and sometimes it becomes a list with recommendation rankings for multiple business partners. The same applies hereafter). It is an object of the present invention to provide a recommendation device and method that can ask a selector to explain the reason when making a selection that is not recommended.

An embodiment for achieving the above object is a recommendation device and method including an open data collection section, a business rule conversion section, an event collection section, an event evaluation section, and a recommendation processing section.

The open data collection unit collects as open data information that can be released or obtained about potential business partners with which the company has already done business, as well as companies with which it can do business in the future, and registers it in the corporate master. In addition to the provisions of laws and regulations, the Business Rules Conversion Department collects compliance policies, ESG policies, sales and purchasing policies, policies, regulations, etc. set by companies themselves, formulates the collected information, and registers it in the business rule master. do. The event collection department collects information on past transaction results with candidate companies, as well as transaction results and evaluation information of companies of the same type as the company, or self-evaluation information of candidate companies as necessary. refers to the business rule master and registers the company's evaluation results for the business partner company in the evaluation master for each individual evaluation axis. When the recommendation processing unit receives a recommendation request regarding a business partner candidate company during the process from internal application to approval of individual transactions within the company, it can recommend the company as a business partner by referring to the corporate master, evaluation master, and business rule master. Present candidate companies and, if necessary, precautions and recommended reasons for doing business with them.

According to the present invention, in the process of selecting a business partner on an existing system that a user uses to carry out business, an overview of the transaction such as the details of the transaction, the transaction size, the timing of the transaction, and the delivery/work site is provided. By entering the terms and conditions regarding the transaction and sending the request as parameters, you can obtain the business partners recommended under your company's business rules and, if necessary, the precautions and reasons for recommendation for each business partner as a response. Furthermore, since there is no need to equip each existing system with a business rule calculation function, it is possible to select business partners based on business rules that are uniformly managed within the company.

1 is a configuration diagram showing a recommendation system to which an embodiment of a recommendation device according to the present invention is applied. 2 is a configuration diagram showing an example of a corporate master that constitutes the recommendation device shown in FIG. 1. FIG. FIG. 3 is an explanatory diagram showing records of the corporate master shown in FIG. 2; FIG. 3 is an explanatory diagram showing a list display presented by the corporate recommendation department. 2 is a flowchart showing a processing procedure for "stable supply of parts" in an embodiment of the recommendation device according to the present invention. It is a flowchart showing the processing procedure of "partner strategy" in one embodiment of the recommendation device according to the present invention. 2 is a flowchart showing a processing procedure for "CSR procurement" or "ESG procurement" in an embodiment of the recommendation method according to the present invention. 2 is a flowchart showing a processing procedure for "changing payment base" in an embodiment of the recommendation device according to the present invention. 2 is a flowchart showing a processing procedure for "legal compliance" in an embodiment of the recommendation method according to the present invention. It is a flowchart showing the processing procedure of "credit limit review" in one embodiment of the recommendation device according to the present invention. It is an explanatory diagram showing an example of risk evaluation by a third party organization. It is an explanatory diagram showing an example of evaluation based on order amount. It is an explanatory diagram showing an example of evaluation based on compliance with laws and regulations. FIG. 2 is an explanatory diagram showing a recommendation/non-recommendation list according to company standards. It is an explanatory diagram showing an example of an independent evaluation list in the own department based on evaluation axes that are not covered by policies and regulations. It is an explanatory diagram showing an example of other department evaluation in the same kind of transaction. FIG. 3 is an explanatory diagram showing a list of normalized values of each axis and comprehensive evaluation values. It is an explanatory diagram showing a list of corporations in descending order of overall evaluation. It is an explanatory diagram showing the optimal evaluation value of each axis necessary for the target corporation to reach the recommended passing value. As a specific example of the optimization shown in FIG. 15, it is an explanatory diagram showing the result of minimizing the least square sum of the number of points of each axis that the target corporation should improve. This is a radar chart showing the evaluation of each corporation using six axes of evaluation data. This is a radar chart that shows the evaluation of each corporation overlappingly.

《Configuration of embodiment》
FIG. 1 is a block diagram showing the configuration of a recommendation system to which an embodiment of a recommendation device according to the present invention is applied.

The recommendation system shown in FIG. 1 includes a recommendation device 10 and an existing system 20 connected to the recommendation device 10 via a network.

The recommendation device 10 is positioned as a so-called rest server, and includes an evaluation master 11, a business rule master 12, and a corporate master 13 as various masters. The recommendation device 10 also includes an event information collection section 14 and an event evaluation section 15 connected to the event collection section 14, the evaluation master 11, and the business rule master 12. It further includes a business rule conversion section 16 connected to the business rule master 12 and an open data collection section 17 connected to the corporate master 13. Furthermore, it includes an evaluation master 11, a business rule master 12, a corporate recommendation unit 18 connected to a corporate master 13 and an existing system 20.

The evaluation master 11 consists of performance information collected from the existing system 20, evaluation lists collected independently by the company, self-evaluation lists from business partners, evaluation lists set independently by the company, and the like.

The performance information collected from the existing system 20 is, for example, records that have been aggregated using an aggregation method (aggregate unit and aggregate value) defined by business rules.

Examples of evaluation lists independently collected by companies include a third-party supplier rating P, an ESG negative list, a suspension/recommendation list from public institutions, and a competitor list.

The self-evaluation list from business partners consists of the "area/item" that the business partner wishes to do business with, the certification list that the business partner has acquired, etc. An example of the evaluation list independently set by a company is the credit limit for each business partner.

The business rule master 12 is necessary to create the evaluation master 11, including the evaluator correction formula, evaluation axis weights, and performance aggregation method, as well as to perform recommended calculations according to requested parameters. , rule information (the data range of the evaluation master 11 targeted by each rule, the evaluation axis weight to be adopted, precautions when selecting a non-recommended business partner, etc.), and each rule (transaction distribution rules, CSR (Corporate Social Calculation methods for social responsibility (social responsibility) or ESG (hereinafter collectively referred to as CSR) compliance rules, invoice payment permission rules, credit limit review standards, etc.) are set.

The evaluator correction formula is a correction formula set for each evaluator. Here, the correction formula is a process of normalizing the evaluation criteria using the correction formula, since there are variations in the individual evaluation criteria. For example, when the correction formula is composed of an activation function and weighting for the result, the activation function is a linear function, a sigmoid function, etc., and the sensitivity that determines the evaluation scale specific to the evaluator and whether it is recommended or not recommended. Rescale around the band (eg -1 to 1). Then, the weighting is set as a coefficient (reliability coefficient, for example, 0 to 1) indicating the reliability of the scores P of a plurality of evaluators, and the final score P is calculated. The weighting coefficients are determined in advance by each evaluator within the company. Alternatively, it is also possible to correct the weighting coefficients by machine learning by comparing the evaluation values before weighting and the results of corporations actually selected as business partners in the past. To give a concrete example of a very simple correction formula, if a certain evaluator gives a five-point rating from 1 to 5 (1 is negative, 3 is neutral, and 5 is positive), the score P is normalized to -1 to 1. Therefore, subtract 3 from the rating P and divide by 2. Furthermore, assess the reliability of the evaluator (for example, let it be 0.5), and then multiply the normalized result by the reliability coefficient (0.5). Correct the score P.

In addition, evaluation of business partners is not uniform, and the evaluation axes and their importance vary depending on the type, scale, and item of the transaction. , it is possible to evaluate business partners corresponding to the transaction details. Here, the evaluation axis weight is a numerical value indicating the importance of the evaluation axis, and the evaluation axis with high importance is set to have a high numerical value.

The performance aggregation method is set such that, for example, aggregation period = past year, aggregation unit = item + business partner, and aggregation value = order amount (million yen).

Transaction distribution rules set the share of each business partner for the same aggregation method with upper and lower tolerance limits, and partner strategy rules set the transaction amount for each business partner for the same aggregation method with upper and lower tolerance limits. This is a standard that sets standards for transaction volumes.

CSR procurement rules define CSR procurement standards such as multipliers, capital ratios, sales ratios, etc. for each transaction type, scale, item, and evaluator. CSR procurement refers to all activities that aim for sound procurement, taking into account various factors such as human rights, labor, and the environment. CSR procurement also includes “ESG procurement.” Note that ESG procurement refers to procurement activities that take into consideration the environment, society, and governance.

Legal compliance rules specify capital (lower limit), precautions, etc. for each transaction type, item, location (country), and delivery destination (country).

The invoice payment possibility rule sets the annual payment amount upper limit for each transaction type, item, budget, and business partner.

The credit limit review rules are based on the rate of change in stock prices, the rate of increase/decrease in capital, the rate of increase/decrease in sales, the rate of increase/decrease in employees, the rate of increase/decrease in gross profit, etc.

The corporate master 13 sets information on corporations that are business partners based on registration information, securities report information, stock price information, and the like.

FIG. 2 shows an example of the corporate master 13. Here, the registration information is information such as a corporate number, corporate name, and location. Information on capital, number of employees, sales, gross profit, and ordinary income is obtained from the securities report information. Further, as stock price information, information such as stock code and N-month fluctuation rate is obtained. Furthermore, FIG. 3 shows the correspondence between each corporate number and the national standard industrial classification code and the international standard industrial classification code.

The event information collection unit 14 collects performance information and the like from the existing system 20, as well as evaluation lists independently collected by companies, self-evaluation lists from business partners, evaluation lists independently set by companies, and the like.

The event evaluation section 15 evaluates the collected information collected by the event information collection section 14 and stores it in the evaluation master 11.

The business rule conversion unit 16 uses a calculation method for the information collected by the event evaluation unit 15 (at least one of a compliance policy, an ESG policy, a sales or purchasing policy, a policy, or a regulation established by the company itself), and a calculation method for the information collected by the event evaluation unit 15. The final evaluation value calculation method to be implemented is stored in the business rule master 12.

The open data collection unit 17 collects public data such as registration information, securities report information, stock price information, etc., and stores it in the corporate master 13.

The corporate recommendation unit 18 is connected to the evaluation master 11, the business rule master 12, the corporate master 13, and the existing system 20, and presents optimal recommendation data to the existing system 20 in response to a recommendation request from the existing system 20. . For example, as a recommendation, as shown in FIG. 4, recommended corporate names are displayed in a list along with evaluation values.

The existing system 20 includes a credit limit management system 21, an RFx management system 22, an order management system 23, an inventory management system 24, a credit and debt management system 25, and the like.

The credit limit management system 21 receives a request to review the credit limit from the corporate recommendation department and resets the credit limit in the evaluation master 11 as an independent evaluation of the company.

The RFx management system 22 manages three request forms that are important when deciding on a supplier, such as a request for information (RFI), a request for proposal (RFP), and a request for quotation (RFQ).

When selecting a supplier, the order management system 23 calls the recommendation device 10 and causes the recommendation device 10 to present highly rated suppliers.

The inventory management system 24 manages and grasps the inventory status using the system, and manages the inventory status using the system to prevent surplus inventory or stockouts.

The credit and debt management system 25 collectively manages information such as balances and payment statuses regarding credits and debts generated in corporate activities.

《Processing procedure of embodiment》
Next, the processing procedure of the recommendation device according to the embodiment of the present invention will be explained based on the flowcharts of FIGS. 5 to 9.

Below, we will focus on the following six points: (1) Stable supply of parts, (2) Partner strategy, (3) CSR procurement, (4) Change in payment base, (5) Legal compliance, and (6) Review of credit limits. The processing will be explained in order. Note that the six processes do not assume a single process, but can be performed in parallel and recommendations can be made by combining the results.

<About stable supply of parts (1)>
With reference to the flowchart of FIG. 5, a processing procedure for stably procuring parts when procuring a procurement target, for example, parts will be described.

In order to secure a stable supply of parts, it is better to strategically disperse parts suppliers rather than limiting them to one location. For example, let us assume that we have decided on a procurement policy that divides the suppliers into three companies: Company A, Company B, and other companies, and sets their respective shares at 5:4:1.

When the order management system 23 requests the corporate recommendation department 18 to select (recommend) a supplier by specifying business rules and parameters (transaction type/item, planned order amount, target budget) (step S1 YES), The corporate recommendation unit 18 refers to the business rule master 12 and selects a transaction distribution rule that the actual order should comply with (parameters match) (step S2).

The latest event evaluation value (order amount aggregated by transaction type, item, budget, and corporation) necessary for calculation of the selected transaction distribution rule is collected from the evaluation master 11 (step S3). Next, the final evaluation value (-1: negative ~ 0: neutral ~ +1: positive) of each corporation is calculated after taking into account the planned amount of the actual order (step S4).

The corporate recommendation unit 18 displays a list of candidate suppliers in descending order of final evaluation value (step S5). If a business partner whose final evaluation value is negative (minus value) is selected from the display list (step S6 YES), the precautions set in the business rule master 12 are displayed as a warning (step S7).

If a caution has been issued, a warning history is stored in the evaluation master 11 (step S8).

The event information collection unit 14 always collects performance data from the existing system 20 in order to update the event evaluation value, and the event evaluation unit 15 collects the event evaluation value (orders aggregated by transaction type, item, budget, and corporation). amount) is updated (step S9).

The business rule master 12 contains, as transaction distribution rules, order lot size, target share (%), tolerance range (upper limit %, lower limit %), for each parameter combination pattern (transaction type, item, budget, corporation). Also, set the calculation method for the final evaluation value (the degree of deviation from the target share is quantified from -1 to +1).

<About (2) partner strategy>
The processing procedure of the partner strategy will be explained with reference to the flowchart of FIG.

As part of their partner strategy, companies have a desire to maintain a certain level of transactions with strategic partner companies every year.

When the order management system 23 specifies business rules and parameters (transaction type, scheduled order amount) and requests the corporate recommendation department 18 to select (recommend) a supplier (step S11 YES), the corporate recommendation department 18 With reference to the rule master 12, a partner strategy rule that the actual order should comply with (parameters match) is selected (step S12).

The latest event evaluation value (order amount aggregated by transaction type and corporation) necessary for calculation of the selected partner strategy rule is collected from the evaluation master 11 (step S13). Next, the final evaluation value (-1: negative ~ 0: neutral ~ +1: positive) of each corporation is calculated after taking into account the planned amount of the actual order (step S14).

The corporate recommendation unit 18 displays a list of candidate suppliers in descending order of final evaluation value (step S15). If a business partner whose final evaluation value is negative (minus value) is selected from the display list (step S16 YES), a warning set in the business rule master 12 is displayed as a warning (step S17).

If a caution has been issued, a warning history is stored in the evaluation master 11 (step S18).

The event information collection unit 14 always collects performance data from the existing system 20 in order to update the event evaluation value, and the event evaluation unit 15 updates the event evaluation value (order amount aggregated by transaction type and corporation). (Step S19).

In addition, the business rule master 12 contains, as partner strategy rules, the order lot size, target order amount, tolerance range (upper limit %, lower limit %), and final evaluation value (target Set the calculation method for the degree of deviation from the order amount (quantified from -1 to +1).

<About (3) CSR procurement>
The processing procedure of CSR procurement will be explained with reference to the flowchart of FIG.

There is a demand for companies to thoroughly implement CSR procurement for direct materials. Here, CSR procurement refers to setting standards from the perspective of social responsibility and CSR when companies select suppliers and set procurement conditions. Through CSR procurement, for example, it is possible to eliminate procurement of products subject to export restrictions, procurement from anti-social companies, etc.

When the RFx management system 22 specifies business rules and parameters (transaction type/item, area, expected amount) and requests the corporate recommendation department 18 to select (recommend) a quotation destination (step S21 YES), the corporate The recommendation unit 18 refers to the business rule master 12 and selects a CSR procurement rule that this request for quotation should comply with (parameters match) (step S22).

The latest event evaluation value (score P by evaluator/company, capital, sales, location, order amount aggregated by transaction type/item/area/company) required for calculation of the selected CSR procurement rule. The information is collected from the evaluation master 11 (step S23). Next, the final evaluation value (-1: negative ~ 0: neutral ~ +1: positive) of each corporation is calculated after considering the scheduled amount of the actual quotation request (step S24).

The corporate recommendation unit 18 displays a list of candidate suppliers in descending order of final evaluation value (step S25). However, the initial display can be narrowed down by items and areas declared by business partners, and the items and area conditions can be changed by operating the screen, or corporations with no transaction history can be targeted for display. If a business partner whose final evaluation value is negative (minus value) is selected from the list display (step S26 YES), the precautions set in the business rule master 12 are displayed as a warning (step S27).

If a caution has been issued, a warning history is stored in the evaluation master 11 (step S28).

The event information collection unit 14 always collects performance data from the existing system 20 in order to update the event evaluation value, and the event evaluation unit 15 collects the event evaluation value (orders aggregated by transaction type, item, area, and corporation). amount) is updated (step S29).

In the business rule master 12, as CSR procurement standards, evaluator correction formulas and evaluation axis weights are set for each parameter combination pattern (for each transaction type, item, scale, and evaluator).

The open data collection unit 17 collects the latest corporate information from stock prices, securities reports, etc. at a fixed cycle (for example, once a day) and updates the corporate master 13.

The event information collection unit 14 collects information in the corporate master 13 (capital , sales, location), and the event evaluation unit 15 updates event evaluation values (score P by evaluator/company, capital, sales, location).

<Regarding the change in payment base in (4)>
When it comes to transactions, there is a desire to switch from invoice-based payments to order-based payments as a measure to strengthen control. In general, invoice-based payment refers to a form in which payment is made as a one-time transaction without registering the business partner as a master on the system as a continuing business partner. Order-based transactions refer to a form in which master registration is performed on the system as a continuing business partner and orders are placed as a designated supplier. In terms of control level, payment on an order basis is a few steps higher, and from the standpoint of a company's internal controls, it is desirable for business partners with whom they have repeated transactions to conduct transactions on an order basis. The following mechanisms will help strengthen this control.

In the flowchart of FIG. 8, the receivables and payables management system 25 requests the corporate recommendation department 18 to select (confirm) a business partner by specifying business rules/parameters (transaction type/item, billing amount, budget) and a target business partner. If so (step S31 YES), the corporate recommendation unit 18 refers to the business rule master 12 and selects a bill payment possible rule that this bill payment should comply with (parameters match) (step S32).

The latest event evaluation values (invoice amounts aggregated by transaction type, item, budget, and corporation) necessary for calculation of the selected invoice payable rule are collected from the evaluation master 11 (step S33). Next, the final evaluation value (-1: negative to 0: neutral to +1: positive) of the target business partner is calculated by adding the billed amount for the actual invoice payment (step S34).

When the final evaluation value is negative (minus value) (step S35 YES), the corporate recommendation unit 18 displays a warning set in the business rule master 12 (step S36).

As a warning display, for example, from the next time onwards, we will encourage you to register your business partners and conduct order-based transactions, and at the same time, we will also display the warning history left in the evaluation master 11 as inappropriate invoice payment transactions. The warning history is stored in the evaluation master 11 (step S37).

The event information collection unit 14 always collects performance data from the existing system 20 in order to update event evaluation values, and the event evaluation unit 15 collects event evaluation values (aggregated by transaction type, item, budget, and corporation). (invoiced amount).

The business rule master 12 contains, as invoice payment possible rules, the invoice payment upper limit amount for each parameter combination pattern (transaction type, item, budget, corporation), and the final evaluation value (the degree of deviation from the invoice amount upper limit). Set the calculation method (expressed as a number from 1 to +1).

<Regarding legal compliance in (5)>
The legal compliance procedure will be described with reference to the flowchart in FIG.
There are many laws and social rules that companies must follow, and violating them will result in a loss of trust in society and result in significant economic loss. , there is a desire to raise awareness of matters that should be taken into consideration in such transactions.

The order management system 23 requests the corporate recommendation department 18 to select (confirm) a supplier by specifying business rules/parameters (transaction type/item, planned order amount, delivery destination (country)) and target business partner. (Step S41 YES), the corporate recommendation unit 18 refers to the business rule master 12 and selects a legal compliance rule that the actual order should comply with (parameters match) (Step S42).

Collect the latest event evaluation values (capital of the target corporation, location (country), location (country) of the parent company, and rating P by evaluator) necessary for calculation of the selected legal compliance rule from the evaluation master 11 ( Step S43). Next, the final evaluation value (-1: negative to 0: neutral to +1: positive) of each target business partner is calculated and displayed (step S44).

If the final evaluation value is negative (step S45 YES), the corporate recommendation unit 18 displays a warning about the precautions in the business rule master 12 corresponding to the selected law (step S46). The history of warning displays is stored in the evaluation master 11 (step S47).

The event information collection unit 14 collects various evaluation information, declaration information, and the latest data from the corporate master 13 in order to constantly update the event evaluation value. Another score P, capital of each corporation, location, parent company) are updated (step S48).

In addition, the business rule master 12 contains capital (lower limit value), precautions, and final evaluation value for each parameter combination pattern (by transaction type, item, location (country), and delivery destination (country)) as legal compliance rules. Set the calculation method (numerically expressed as a legal warning level of -1 to +1).

<Regarding the review of the credit limit in (6)>
The procedure of the credit limit review process will be described with reference to the flowchart of FIG. 10.

There is a desire among companies to review their credit limits in a timely manner.
The credit limit management system 21 requests the corporate recommendation unit 18 to select (confirm) a business partner for which the credit limit is applicable by specifying business rules/parameters (transaction period, payment site, transaction amount) and target business partner. Then (step S51 YES), the corporate recommendation unit 18 refers to the evaluation master 11 and selects a credit limit review rule that this transaction should comply with (with which the parameters match). (Step S52).

The latest event evaluation values required to calculate the selected credit limit review rule (total amount of receivables balance and uninspected amount, debt balance, score P by evaluator, capital, sales, stock price fluctuation rate, capital The rate of increase/decrease in money, the rate of increase/decrease in sales, the rate of increase/decrease in the number of employees, the rate of increase/decrease in gross profit) are collected from the evaluation master 11 (step S53). Next, the final evaluation value (-1: negative to 0: neutral to +1: positive) of the target corporation is calculated by adding the transaction amount of the main credit limit check (step S54).

If the final evaluation value is negative (step S55 YES), the corporate recommendation unit 18 displays a warning set in the business rule master 12 (step S56).

The event information collection unit 14 collects performance data from the existing system 20 in order to update the event evaluation value, and the event evaluation unit 15 collects the event evaluation value (the total amount of outstanding receivables and uninspected amounts aggregated by corporation). , debt balance) is updated (step S57).

In addition, the event information collection unit 14 collects information from the corporate master 13 in addition to the latest score P from within the company or a third-party organization, and self-reports from business partners (capital, sales, number of employees, gross profit). The event evaluation unit 15 collects information (capital, sales, stock price fluctuation rate, increase/decrease rate of capital, increase/decrease rate of sales, increase/decrease rate of number of employees, increase/decrease rate of gross profit), and calculates the event evaluation value ( Update score P, capital, sales) by evaluator/company.

In addition, Business Rule Master 12 contains the credit limit calculation method and the credit limit review criteria (stock price fluctuation rate "±30%", capital increase/decrease rate "±50%", sales increase/decrease rate "±20%"). ”, employee increase/decrease rate “±10%”, gross profit increase/decrease rate “±10%”, etc.).

The open data collection unit 17 collects the latest corporate information from stock prices, securities reports, etc. at a fixed cycle (for example, once a day) and updates the corporate master 13.

The event evaluation unit 15 refers to the credit limit review rules in the business rule master 12 for the record, and uses review criteria (stock price fluctuation rate, capital increase/decrease rate, sales increase/decrease rate, increase/decrease rate in the number of employees). , rate of increase/decrease in gross profit) is output as a list and notified to the administrator as a target whose in-house score P needs to be reviewed (step S58).

<Specific example of evaluation value calculation and display example>
Next, a specific display example of recommended corporations realized by the recommendation device according to the present invention will be explained based on FIGS. 11 to 22.

In this example, comprehensive evaluation is performed on six axes, Axis A to Axis F. Further, for each axis A to F, a normalization method for the elements composing the axis is set. At this time, the evaluator correction formula is set as a calculation formula using the score P of each component as a variable. For example, if the activation function (P-50)/50 is determined that the evaluator's rating P linearly expresses the recommendation priority, then the activation function type is a linear function, and the evaluator's scale is shows a calculation formula for a maximum value of 100 points, a minimum value of 0 points, and a neutral value of 50 points. It also indicates that the evaluator is given a weight (1.0) as a reliable institution.

Axis A aggregates the results obtained using the evaluator correction formula from the ratings P of three third-party organizations x, y, and z.

Axis B aggregates the results obtained using the evaluator correction formula from the scores P of the three trading policies e, f, and g.

Axis C aggregates the results obtained using the evaluator correction formula from the scores P for the three legal responses h, i, and j.

Axis D is whether the company is covered by the list collected and created by the company (negative list of companies with which transactions are not recommended, list of companies that have made an ESG declaration, list of companies that have concluded a valid basic contract). ``1'' for non-target corporations, ``0'' for non-target corporations), and the results obtained using the evaluator correction formula are totaled.

Axis E aggregates the results obtained using the evaluator correction formula from the scores P of the three evaluation items of the own department's own evaluations u, v, and w.

Axis F aggregates the results obtained using the evaluator correction formula from the scores P on evaluation axis E for other departments r, s, and t in the same type of transaction.

FIG. 11 shows an example of axis A evaluation values, which are risk evaluations by a third-party organization. FIG. 12 shows axis B evaluation values, which are evaluation examples based on order amounts. FIG. 13 shows axis C evaluation values as an example of evaluation based on legal compliance. FIG. 14 shows axis D evaluation values, which are a recommendation/non-recommendation list based on company standards. FIG. 15 shows axis E evaluation values, which are an example of the own department's own evaluation list on evaluation axes that are not covered by policies and regulations. FIG. 16 shows axis F evaluation values, which are an example of independent evaluations of other departments in similar transactions.

Then, as shown in FIG. 17, the evaluation value for each axis A to F of each corporation a to f is multiplied by the evaluation axis weight of the business rule to obtain a comprehensive evaluation value. Next, as shown in FIG. 18, the evaluation results shown in FIG. 17 are sorted and displayed in descending order of overall evaluation.

In addition, as shown in Figure 19, target corporation a shows the optimal evaluation value for each axis necessary to reach the recommended passing value (for example, 0.5), and what improvements should be made to conduct business with non-recommended corporations. Show what you should do.

In this case, as a highly feasible improvement measure, in order to minimize the deviation from the current score P, as shown in Figure 20, the number of points in each axis that the target corporation must improve and , the combination that minimizes the least square sum is calculated and displayed.

As a display example, the evaluation results of each corporation a to f are displayed in a radar chart as shown in FIG. Furthermore, as shown in FIG. 22, the evaluation results of each corporation a to f may be displayed in a superimposed manner on one radar chart.

Although the embodiments of the present invention have been described above, this is merely an example and does not limit the scope of the invention. Various omissions, substitutions, and changes can be made to the embodiments without departing from the gist of the invention.

10... Recommendation device, 11... Evaluation master, 12... Business rule master, 13... Corporate master, 14... Event information collection section, 15... Event evaluation section, 16... Business rule conversion section, 17... Open data collection section, 18 ...Corporate recommendation department (recommendation processing department), 20...Existing system, 21...Credit limit management system, 22...RFx management system, 23...Order management system, 24...Inventory management system, 25...Receivables and payables management system

Claims (10)

An open data collection department that collects publicly available or sold corporate information regarding potential business partners as open data and registers it in the corporate master;
Business rules that collect at least one of the compliance policies, ESG policies, sales and purchasing policies, policies, and regulations that companies set themselves in addition to the provisions of laws and regulations, formulate each collected information, and register it in the business rule master. A conversion section;
an event evaluation unit that refers to past transaction performance information with the business partner candidate company and the business rule master to register the company's evaluation results for the business partner candidate company in the evaluation master for each individual evaluation axis;
and a recommendation processing unit that refers to the corporate master, evaluation master, and business rule master to present companies that can be recommended as business partners when there is a recommendation request regarding business partners;
The recommendation processing unit is a recommendation device that recommends optimal business partner companies according to business rules that differ from company to company. The apparatus according to claim 1, comprising:
In addition to internal evaluation information, the evaluation master collects external evaluations that include at least one of the transaction records and evaluation information of companies of the same type as the company, or self-evaluation information of potential business partners, and stores each evaluation information. are expressed on the same evaluation scale, and the evaluation values normalized by the evaluator correction formula multiplied by the reliability coefficient of each evaluator are registered in the evaluation master, and the recommendation processing unit uses the evaluation values based on the evaluation values. , a recommendation device that presents companies that can be recommended as business partners. The device according to claim 1 or 2,
When recommending business partners, the recommendation processing unit does not recommend a single company, but creates a list of recommended companies with which business can be done, and displays the recommended business partner companies in descending order of evaluation value. Device. The device according to any one of claims 1 to 3,
The recommendation processing unit is a recommendation device that displays, for each company, the reason for recommending the business partner candidate or precautions when doing business with the business partner candidate. The device according to any one of claims 1 to 4,
When the recommendation processing unit selects a business partner that is not recommended, it urges compliance with business rules by calling out precautions for that transaction type, and also requests a description or input of the reason for acting outside the business rules. , Recommendation device. A recommendation method executed on a rest server configured with a computer,
Collect publicly available or sold company information regarding potential business partners as open data, register it in the corporate master,
In addition to the provisions of laws and regulations, companies collect at least one of their own compliance policies, ESG policies, sales and purchasing policies, policies, and regulations, formulate each collected information, and register it in a business rule master.
Register the company's evaluation results for the business partner candidate company in the evaluation master for each individual evaluation axis by referring to past transaction performance information with the business partner candidate company and the business rule master,
When there is a recommendation request regarding a business partner, the company that can be recommended as a business partner is presented by referring to the corporate master, evaluation master, and business rule master. Recommendation method to recommend business partners. The recommendation method according to claim 6,
In addition to internal evaluation information, the evaluation master collects at least one of the transaction records and evaluation information of companies of the same type as the company, or self-evaluation information of potential business partners, including external evaluations. A recommendation method in which companies are registered as standardized evaluation values and based on those evaluation values, companies that can be recommended as business partners are presented. The recommendation method according to claim 6 or 7,
When recommending business partners, instead of recommending a single company, this recommendation method creates a list of recommended companies with which business can be done and displays the recommended business partners in descending order of evaluation value. The recommendation method according to any one of claims 6 to 8,
A recommendation method that displays the reasons for recommending business partner candidates or precautions when doing business with business partner candidates for each company. The recommendation method according to any one of claims 6 to 9,
If an attempt is made to select a business partner that is not recommended, a recommendation method that prompts compliance with business rules by calling out precautions for that transaction type, and requests a description or input of the reason for the action outside of business rules. .

Images