Multi-field confidence data international trade enterprise credit evaluation algorithm

Abstract

The invention relates to the technical field of credit evaluation, and discloses a multi-field confidence data international trade enterprise credit evaluation algorithm, and the algorithm comprises the steps: collecting multi-field confidence data, preprocessing the multi-field confidence data, selecting an evaluation feature set, and carrying out credit evaluation. According to the method, by collecting the multi-field confidence data, the defect that an existing method depends on single-field data is overcome, comprehensiveness of credit evaluation is achieved, multi-field data are preprocessed, data quality and usability are improved, and a foundation is laid for subsequent accurate evaluation. The evaluation feature set is determined by using the feature selection algorithm, the influence degree of data in each field on credit evaluation is accurately measured, the accuracy is improved, and the multi-factor dynamic credit evaluation algorithm performs evaluation based on the dynamic weight set and the evaluation feature set, so that the weight is adjusted in time based on the international trade environment change and the enterprise characteristics, and the credit evaluation efficiency is improved. And the adaptability and the efficiency are enhanced.

Description

Technical Field

[0001] The present application relates to the technical field of credit evaluation, and specifically to a credit evaluation algorithm for international trade enterprises using multi-domain trust data. Background Art

[0002] Existing credit evaluation methods for international trade enterprises often rely on data from a single field, such as financial data or market performance, and ignore the comprehensive impact of data from multiple fields. In addition, existing methods are inefficient in processing large-scale heterogeneous data and are difficult to adapt to the dynamically changing international trade environment.

[0003] In summary, there is an urgent need for a new technical solution that uses multi-field confidence data to conduct credit evaluation of international trade enterprises, so as to improve the comprehensiveness, adaptability, accuracy and efficiency of credit evaluation of international trade enterprises. Summary of the invention

[0004] The purpose of this application is to provide a credit evaluation algorithm for international trade enterprises based on multi-domain trust data to solve the technical problems raised in the above-mentioned background technology.

[0005] To achieve the above objectives, the present application discloses the following technical solutions: a credit evaluation algorithm for international trade enterprises based on multi-domain trust data, including:

[0006] S1: Collect multi-field confidence data; wherein the multi-field confidence data is used to characterize the credit data of international trade enterprises in different fields;

[0007] S2: preprocessing the multi-domain confidence data; wherein the preprocessing at least includes data cleaning, data standardization and data normalization processing;

[0008] S3: Selecting an evaluation feature set of the multi-domain confidence data; wherein, when selecting, using a feature selection algorithm pre-trained by machine learning to determine the evaluation feature set of the currently performed credit evaluation, the feature selection algorithm is used to extract corresponding features from the multi-domain confidence data, calculate corresponding credit evaluation impact factors, and output the evaluation feature set after sorting, the credit evaluation impact factor is used to characterize the degree of influence of the confidence data in a certain field on the current credit evaluation, and the evaluation feature set is used to perform credit evaluation;

[0009] S4: Conduct credit evaluation using a preset multi-factor dynamic credit evaluation algorithm; wherein the multi-factor dynamic credit evaluation algorithm is used to conduct credit evaluation based on a preset multi-factor dynamic weight set in combination with the evaluation feature set, wherein the multi-factor dynamic weight set is a set of dynamic weight values ​​obtained based at least on the actual degree of influence of changes in the international trade environment and the company's own characteristics on the credit evaluation, and the multi-factor dynamic weight set is used to adjust the weights of the features in the evaluation feature set in the current credit evaluation.

[0010] Preferably, the multi-field trust data includes at least: text data in the fields of financial data, market performance, legal compliance records, and supply chain stability, real-time collection of dynamic data on the logistics and production links of enterprises, professional credit-related data obtained in cooperation with international credit rating agencies and industry associations, and the use of big data crawler technology to capture unstructured data of news and social media comments related to enterprises from the Internet.

[0011] Preferably, the operation process of the feature selection algorithm includes:

[0012] Analyze the multi-field confidence data to determine a feature range; wherein the feature range is a collection of various features related to credit evaluation that are preliminarily determined from the multi-field confidence data;

[0013] Extract corresponding features from the confidence data in different fields, calculate the importance index of each feature and the credit evaluation target, and define the importance index as the credit evaluation impact factor; the credit evaluation impact factor is used to quantify the importance of the confidence data in a certain field to the current credit evaluation;

[0014] The features are sorted based on the credit evaluation impact factor, and the features with higher rankings are selected based on a preset number of features to obtain the evaluation feature set; wherein the number of features is the number of types of features in the evaluation feature set.

[0015] Preferably, the credit evaluation influencing factor is specifically: calculating the information gain of the feature and the credit evaluation target.

[0016] Preferably, the evaluation feature set is specifically:

[0017] The features that have a higher impact on the current credit rating are screened out, and the features at least include the company's financial ratios, market share change rate, indicators of the goodness of its legal compliance record, and quantitative indicators of supply chain stability.

[0018] Preferably, the operation process of the multi-factor dynamic credit evaluation algorithm includes:

[0019] Determine the initial dynamic weight set of the multi-factors based on changes in the international trade environment and the enterprise's own characteristics;

[0020] Calculate the features in the evaluation feature set and the corresponding weights to obtain a preliminary credit evaluation result;

[0021] The updated multi-factor dynamic weight set is obtained, and the credit evaluation result is updated.

[0022] Preferably, the multi-factor dynamic weight set is specifically:

[0023] The multi-factor dynamic weight set includes a plurality of weight values, and corresponds to each feature value in the evaluation feature set of confidence data in different fields; wherein the feature value is a score obtained by quantifying the feature in the evaluation feature set;

[0024] The determination process of the multi-factor dynamic weight set is:

[0025] A1: Analyze the economic policy stability of major trading countries by collecting and quantifying the adjustment frequency of fiscal and monetary policies and the consistency of policies of trading countries within a preset analysis period;

[0026] A2: Analyze exchange rate fluctuations and calculate the exchange rate fluctuation range of major trading currencies during the analysis period;

[0027] A3: Quantify the size of the enterprise and its industry position;

[0028] A4: Based on the quantification of steps A1 to A4, the multi-factor dynamic weight set is obtained.

[0029] Preferably, the preliminary credit evaluation result is specifically:

[0030] The preliminary credit evaluation result is obtained by calculating each feature value in the evaluation feature set and the corresponding initial multi-factor dynamic weight set.

[0031] Preferably, the obtaining of the updated multi-factor dynamic weight set is specifically:

[0032] When there are updated feature values ​​of the evaluation feature set of confidence data in different fields, steps A1 to A4 are re-executed to obtain an updated multi-factor dynamic weight set.

[0033] Preferably, the updating of the credit evaluation result is specifically as follows:

[0034] Obtaining the updated multi-factor dynamic weight set and each feature value in the updated evaluation feature set of confidence data in different fields;

[0035] Each feature value in the updated evaluation feature set is calculated with the corresponding updated multi-factor dynamic weight set to obtain an updated credit evaluation result.

[0036] Beneficial effects: The credit evaluation algorithm for international trade enterprises based on multi-field confidence data of this application has achieved a comprehensive improvement in the credit evaluation of international trade enterprises by using the credit evaluation algorithm for international trade enterprises based on multi-field confidence data. Among them, by collecting multi-field confidence data, covering multiple fields such as financial data, market performance, legal compliance records, supply chain stability, and real-time collection of enterprise dynamic data, cooperation with professional institutions to obtain data and capture unstructured data on the Internet, it overcomes the defects of existing methods that rely on data in a single field, achieves the comprehensiveness of credit evaluation, pre-processes multi-field data, improves data quality and availability, and lays the foundation for subsequent accurate evaluation. The evaluation feature set is determined by using the feature selection algorithm to accurately measure the degree of influence of data in various fields on credit evaluation, thereby improving accuracy. In addition, the multi-factor dynamic credit evaluation algorithm is based on the dynamic weight set combined with the evaluation feature set for evaluation, which realizes timely adjustment of weights based on changes in the international trade environment and the characteristics of the enterprise itself, and enhances adaptability and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 A flowchart of the credit evaluation algorithm for international trade enterprises based on multi-domain confidence data provided in the embodiment of the present application. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present application are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

[0040] In this article, the term "comprising" is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of more restrictions, the elements defined by the sentence "comprising..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0041] This embodiment discloses Figure 1 The multi-domain trust data international trade enterprise credit evaluation algorithm shown includes:

[0042] S1: Collect multi-field confidence data; the multi-field confidence data is used to characterize the credit data of international trade enterprises in different fields;

[0043] S2: Preprocessing multi-domain confidence data; wherein the preprocessing at least includes data cleaning, data standardization and data normalization;

[0044] S3: Selecting an evaluation feature set of multi-domain confidence data; wherein, when selecting, using a feature selection algorithm pre-trained by machine learning to determine the evaluation feature set of the current credit evaluation, the feature selection algorithm is used to extract corresponding features from the multi-domain confidence data, calculate the corresponding credit evaluation impact factors, sort them, and then output the evaluation feature set, the credit evaluation impact factor is used to characterize the degree of influence of the confidence data in a certain field on the current credit evaluation, and the evaluation feature set is used to perform credit evaluation;

[0045] S4: Use a preset multi-factor dynamic credit evaluation algorithm to conduct credit evaluation; wherein the multi-factor dynamic credit evaluation algorithm is used to conduct credit evaluation based on a preset multi-factor dynamic weight set combined with an evaluation feature set, and the multi-factor dynamic weight set is a set of dynamic weight values ​​obtained based on at least the actual degree of influence of changes in the international trade environment and the company's own characteristics on the credit evaluation, and the multi-factor dynamic weight set is used to adjust the weights of the features in the evaluation feature set in the current credit evaluation.

[0046] In a specific application, this embodiment uses multi-field data collection, preprocessing, feature selection, and multi-factor dynamic credit evaluation technologies to achieve a comprehensive and accurate assessment of the credit of international trade enterprises. For example, for a company engaged in the import and export of electronic products, first collect its financial data (such as balance sheets, income statements, etc.), market performance (product market share, customer satisfaction, etc.), legal compliance records (whether there are violations and penalties, etc.), supply chain stability (supplier reliability, etc.) and other multi-field data. Then perform preprocessing such as data cleaning, determine key features through feature selection algorithms, and finally use multi-factor dynamic credit evaluation algorithms to obtain credit evaluation results.

[0047] Based on the above, this embodiment uses the credit evaluation algorithm of international trade enterprises with multi-field confidence data to achieve a comprehensive improvement in the credit evaluation of international trade enterprises. Among them, by collecting multi-field confidence data, covering multiple fields such as financial data, market performance, legal compliance records, supply chain stability, and real-time collection of enterprise dynamic data, cooperating with professional institutions to obtain data and crawling Internet unstructured data, it overcomes the defects of existing methods that rely on single-field data, achieves the comprehensiveness of credit evaluation, pre-processes multi-field data, improves data quality and availability, and lays the foundation for subsequent accurate evaluation. The evaluation feature set is determined by the feature selection algorithm, and the degree of influence of data in various fields on credit evaluation is accurately measured, which improves accuracy. The multi-factor dynamic credit evaluation algorithm is based on the dynamic weight set combined with the evaluation feature set for evaluation, which realizes the timely adjustment of weights based on changes in the international trade environment and the characteristics of the enterprise itself, and enhances adaptability and efficiency.

[0048] Specifically, multi-field trust data includes at least: text data in the fields of financial data, market performance, legal compliance records, and supply chain stability; real-time collection of dynamic data on the company's logistics and production links; professional credit-related data obtained in cooperation with international credit rating agencies and industry associations; and the use of big data crawler technology to capture unstructured data on company-related news and social media comments from the Internet.

[0049] It should be noted that this embodiment uses existing data collection technology to achieve the collection of multi-domain confidence data. In a specific application, this embodiment uses multiple data collection channel technologies to achieve comprehensive data acquisition. For example, for the same electronic product import and export company, financial data such as sales and profits in the past year are obtained from the company's financial system; market performance data such as market share changes in different regions are obtained through market research companies; government departments and industry databases are queried to obtain legal compliance records; suppliers are communicated to obtain supply chain stability information; sensors are used to collect real-time data such as transportation time in the logistics link; professional credit assessment data is obtained in cooperation with international credit rating agencies; and product evaluation and reputation information about the company is captured from news websites and social media using web crawlers.

[0050] Based on the above, this embodiment utilizes multi-channel data collection to achieve comprehensive coverage of credit data of international trade enterprises and improve the accuracy and reliability of evaluation.

[0051] Specifically, the operation process of the feature selection algorithm includes:

[0052] Analyze the multi-domain confidence data to determine the feature range; wherein the feature range is a collection of various features related to credit evaluation that are preliminarily determined from the multi-domain confidence data;

[0053] Corresponding features are extracted from the confidence data in different fields, and the importance index of each feature and the credit evaluation target is calculated, and the importance index is defined as the credit evaluation impact factor; the credit evaluation impact factor is used to quantify the importance of the confidence data in a certain field to the current credit evaluation;

[0054] The features are ranked based on the credit evaluation influencing factors, and the features with higher rankings are selected based on the preset number of features to obtain an evaluation feature set; wherein the number of features is the number of types of features in the evaluation feature set.

[0055] In a specific application, this embodiment uses feature selection algorithm technology to achieve the screening of key features. For example, for the multi-field data collected from electronic product import and export enterprises, analyze and determine the range of features that may be related to credit evaluation, such as debt-to-asset ratio, current ratio, etc. in financial data, customer satisfaction, market share growth rate, etc. in market performance, calculate the importance index of these features and credit evaluation targets, such as calculating the contribution of each feature to credit evaluation through the existing information gain algorithm, sorting according to the contribution, and selecting several features with higher rankings to form an evaluation feature set, such as debt-to-asset ratio, market share growth rate, supplier stability index, etc.

[0056] Based on the above, this embodiment uses a feature selection algorithm to screen out key features that have a greater impact on credit evaluation from multi-field data, thereby improving the efficiency and accuracy of the evaluation.

[0057] Specifically, the credit evaluation influencing factor is: the information gain between the calculated feature and the credit evaluation target.

[0058] In a specific application, this embodiment uses the existing information gain calculation technology to determine the credit evaluation influencing factors. For example, for electronic product import and export enterprises, the information gain between the asset-liability ratio feature in the financial data and the credit evaluation target is calculated. If the change in the asset-liability ratio can greatly affect the credit evaluation result, then its information gain is high, indicating that the feature is more important to the credit evaluation. By calculating the information gain of each feature, the importance of confidence data in different fields to the current credit evaluation can be quantified.

[0059] Based on the above, this embodiment uses information gain calculation to accurately quantify the importance of each feature to credit evaluation, providing a basis for feature selection and weight allocation.

[0060] Specifically, the evaluation feature set is:

[0061] The features that have a higher impact on the current credit rating are screened out, and the features at least include the company's financial ratios, market share change rate, indicators of the goodness of its legal compliance record, and quantitative indicators of supply chain stability.

[0062] In a specific application, this embodiment uses existing feature screening technology to achieve the determination of the evaluation feature set. For example, for electronic product import and export enterprises, the feature selection algorithm selects the enterprise's financial ratios (such as asset turnover), market share change rate (share growth in the international market in the past six months), goodness of legal compliance records (whether there are major violations of laws and regulations and the number of penalties), and quantitative indicators of supply chain stability (supply stability of major suppliers) and other features to form the evaluation feature set. These features have a high impact on the credit evaluation of the enterprise and can more comprehensively reflect the credit status of the enterprise.

[0063] Based on the above, this embodiment uses feature screening to determine a feature set that has a significant impact on credit evaluation, providing a basis for accurately evaluating corporate credit.

[0064] Specifically, the operation process of the multi-factor dynamic credit evaluation algorithm includes:

[0065] Determine the initial multi-factor dynamic weight set based on changes in the international trade environment and the company's own characteristics;

[0066] Combine the features in the evaluation feature set and the corresponding weights to calculate and obtain a preliminary credit evaluation result;

[0067] Obtain updated multi-factor dynamic weight sets and update credit evaluation results.

[0068] In a specific application, this embodiment uses a multi-factor dynamic credit evaluation algorithm technology to achieve dynamic adjustment of credit evaluation. Exemplarily, for electronic product import and export enterprises, first consider changes in the international trade environment, such as trade policy adjustments, exchange rate fluctuations, etc., as well as the company's own characteristics, such as company size, industry status and other factors to determine the initial multi-factor dynamic weight set. Then the features in the evaluation feature set (such as financial ratios, market share change rates, etc.) are calculated with the corresponding weights to obtain a preliminary credit evaluation result. Over time, when the international trade environment changes (such as a country raising tariffs) or a major event occurs in the company itself (such as the launch of a new product), the weight set is re-determined and the credit evaluation result is updated.

[0069] Based on the above, this embodiment utilizes a multi-factor dynamic credit evaluation algorithm to achieve real-time adjustment of credit evaluation according to the international trade environment and changes in the enterprise itself, thereby improving the accuracy and adaptability of the evaluation.

[0070] Specifically, the multi-factor dynamic weight set is:

[0071] The multi-factor dynamic weight set includes multiple weight values, and corresponds to each feature value in the evaluation feature set of confidence data in different fields; wherein the feature value is a score obtained by quantifying the features in the evaluation feature set;

[0072] The process of determining the multi-factor dynamic weight set is:

[0073] A1: Analyze the economic policy stability of major trading countries by collecting and quantifying the adjustment frequency of fiscal and monetary policies and the consistency of policies of trading countries within a preset analysis period;

[0074] A2: Analyze exchange rate fluctuations and calculate the fluctuation range of exchange rates of major trading currencies during the analysis period;

[0075] A3: Quantify the size of the enterprise and its industry position;

[0076] A4: Based on the quantification of steps A1 to A4, a multi-factor dynamic weight set is obtained.

[0077] In a specific application, this embodiment uses multi-factor quantification technology to achieve the determination of dynamic weight sets. For example, for electronic product import and export companies, the economic policy stability of major trading countries in the past quarter is analyzed, such as by counting the number of fiscal policy adjustments as 1, the number of monetary policy adjustments as 2, and the policy has a certain consistency, and a higher stability score is given. Calculate the exchange rate fluctuations of major trading currencies in the same period, and give a higher exchange rate stability score if the fluctuation is small. At the same time, quantify the scale of the enterprise according to the total assets, operating income, etc. of the enterprise, and quantify the industry position according to market share and brand awareness. Combining the quantitative results of these factors, determine the multi-factor dynamic weight set, such as the financial ratio weight is 30%, the market share change rate weight is 25%, etc.

[0078] Based on the above, this embodiment utilizes multi-factor quantification to achieve the determination of a weight set that can dynamically reflect the international trade environment and the characteristics of the enterprise itself, thereby providing support for accurate credit evaluation.

[0079] The specific, preliminary credit evaluation results are as follows:

[0080] Each feature value in the evaluation feature set is calculated with the corresponding initial multi-factor dynamic weight set to obtain a preliminary credit evaluation result.

[0081] In a specific application, this embodiment uses weighted calculation technology to determine the preliminary credit evaluation results. For example, for electronic product import and export enterprises, assuming that the financial ratio feature value in the evaluation feature set is 80 points (out of 100 points), the corresponding initial weight is 30%; the market share change rate feature value is 75 points, the weight is 25%, etc. Multiply each feature value with the corresponding weight and add them up to obtain a preliminary credit evaluation result, and calculate a comprehensive credit score to preliminarily reflect the credit status of the enterprise in the current environment. It should be noted that the determination of the preliminary credit evaluation results is not limited to the weighted calculation technology used in this embodiment.

[0082] Based on the above, this embodiment uses weighted calculation to obtain a preliminary credit evaluation result by combining characteristic values ​​and weights, providing a basis for subsequent adjustments.

[0083] Specifically, obtain the updated multi-factor dynamic weight set, specifically:

[0084] When there are updated feature values ​​in the evaluation feature set of confidence data in different fields, steps A1 to A4 are re-executed to obtain an updated multi-factor dynamic weight set.

[0085] In a specific application, this embodiment uses the re-execution of A1 to A4 to achieve the update of the v weight set. For example, for electronic product import and export enterprises, when new situations arise, such as changes in characteristic values ​​caused by the expansion of production scale of the enterprise, or major changes in the international trade environment (such as a country suddenly imposing high tariffs), re-execute steps A1 to analyze the stability of economic policies of major trading countries (the stability score may be reduced due to tariff adjustments), A2 to analyze exchange rate fluctuations (tariff adjustments may cause exchange rate fluctuations), A3 to quantify enterprise scale (re-quantify after scale expansion) and industry status, and other steps, thereby obtaining an updated multi-factor dynamic weight set.

[0086] Based on the above, this embodiment utilizes re-quantization to timely update the weight set according to new circumstances, thereby ensuring the accuracy and adaptability of the credit evaluation.

[0087] Specifically, update the credit evaluation results, specifically:

[0088] Obtaining updated multi-factor dynamic weight sets and updated individual feature values ​​in evaluation feature sets of confidence data in different fields;

[0089] Each feature value in the updated evaluation feature set is calculated with the corresponding updated multi-factor dynamic weight set to obtain an updated credit evaluation result.

[0090] In a specific application, this embodiment uses the update calculation technology to achieve the update of the credit evaluation results. For example, for electronic product import and export enterprises, after obtaining the updated multi-factor dynamic weight set and the updated characteristic values ​​in the evaluation characteristic set, such as the new financial ratio characteristic value after the enterprise scale is expanded is 85 points, the new weight is 35%, etc. The updated characteristic value is multiplied and accumulated with the corresponding updated weight to obtain the updated credit evaluation result, reflecting the credit status of the enterprise under the new situation.

[0091] Based on the above, this embodiment uses update calculation to timely update the credit evaluation results according to the new weight set and feature value, providing the latest reference for relevant decisions.

[0092] In summary, the credit evaluation algorithm for international trade enterprises with multi-field confidence data in this embodiment uses the credit evaluation algorithm for international trade enterprises with multi-field confidence data to achieve a comprehensive improvement in the credit evaluation of international trade enterprises. Among them, by collecting multi-field confidence data, covering multiple fields such as financial data, market performance, legal compliance records, supply chain stability, and real-time collection of enterprise dynamic data, cooperation with professional institutions to obtain data and capture Internet unstructured data, it overcomes the defects of existing methods that rely on single-field data, achieves the comprehensiveness of credit evaluation, pre-processes multi-field data, improves data quality and availability, and lays the foundation for subsequent accurate evaluation. The evaluation feature set is determined by using the feature selection algorithm, and the degree of influence of data in various fields on credit evaluation is accurately measured, which improves accuracy. The multi-factor dynamic credit evaluation algorithm is based on the dynamic weight set combined with the evaluation feature set for evaluation, which realizes the timely adjustment of weights based on changes in the international trade environment and the characteristics of the enterprise itself, and enhances adaptability and efficiency.

[0093] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein or their combination. For software implementation, part or all of the flow of the embodiment can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein the communication medium includes any medium that is convenient for transmitting a computer program from one place to another. The storage medium can be any available medium that a computer can access. The computer-readable storage medium can include but is not limited to RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer.

[0094] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A credit evaluation algorithm for international trade enterprises based on multi-domain trust data, characterized in that: include: S1: Collect multi-field confidence data; wherein the multi-field confidence data is used to characterize the credit data of international trade enterprises in different fields; S2: preprocessing the multi-domain confidence data; wherein the preprocessing at least includes data cleaning, data standardization and data normalization processing; S3: Selecting an evaluation feature set of the multi-domain confidence data; wherein, when selecting, using a feature selection algorithm pre-trained by machine learning to determine the evaluation feature set of the currently performed credit evaluation, the feature selection algorithm is used to extract corresponding features from the multi-domain confidence data, calculate corresponding credit evaluation impact factors, and output the evaluation feature set after sorting, the credit evaluation impact factor is used to characterize the degree of influence of the confidence data in a certain field on the current credit evaluation, and the evaluation feature set is used to perform credit evaluation; S4: Conduct credit evaluation using a preset multi-factor dynamic credit evaluation algorithm; wherein the multi-factor dynamic credit evaluation algorithm is used to conduct credit evaluation based on a preset multi-factor dynamic weight set in combination with the evaluation feature set, wherein the multi-factor dynamic weight set is a set of dynamic weight values ​​obtained based at least on the actual degree of influence of changes in the international trade environment and the company's own characteristics on the credit evaluation, and the multi-factor dynamic weight set is used to adjust the weights of the features in the evaluation feature set in the current credit evaluation.

2. The international trade enterprise credit evaluation algorithm according to claim 1 is characterized in that: The multi-field trust data include at least: financial data, market performance, legal compliance records, text data in the field of supply chain stability, real-time collection of dynamic data of enterprises in logistics and production links, professional credit-related data obtained in cooperation with international credit rating agencies and industry associations, and the use of big data crawler technology to capture unstructured data of news and social media comments related to the enterprise from the Internet.

3. The international trade enterprise credit evaluation algorithm according to claim 1 is characterized in that: The operation process of the feature selection algorithm includes: Analyze the multi-field confidence data to determine a feature range; wherein the feature range is a collection of various features related to credit evaluation that are preliminarily determined from the multi-field confidence data; Extract corresponding features from the confidence data in different fields, calculate the importance index of each feature and the credit evaluation target, and define the importance index as the credit evaluation impact factor; the credit evaluation impact factor is used to quantify the importance of the confidence data in a certain field to the current credit evaluation; The features are sorted based on the credit evaluation impact factor, and the features with higher rankings are selected based on a preset number of features to obtain the evaluation feature set; wherein the number of features is the number of types of features in the evaluation feature set.

4. The international trade enterprise credit evaluation algorithm according to claim 3 is characterized in that: Credit evaluation influencing factors, specifically: calculating the information gain between the features and the credit evaluation target.

5. The international trade enterprise credit evaluation algorithm according to claim 3 is characterized in that: The evaluation feature set is specifically: The features that have a higher impact on the current credit rating are screened out, and the features at least include the company's financial ratios, market share change rate, indicators of the goodness of its legal compliance record, and quantitative indicators of supply chain stability.

6. The international trade enterprise credit evaluation algorithm according to claim 1 is characterized in that: The operation process of the multi-factor dynamic credit evaluation algorithm includes: Determine the initial dynamic weight set of the multi-factors based on changes in the international trade environment and the enterprise's own characteristics; Calculate the features in the evaluation feature set and the corresponding weights to obtain a preliminary credit evaluation result; The updated multi-factor dynamic weight set is obtained, and the credit evaluation result is updated.

7. The international trade enterprise credit evaluation algorithm according to claim 6 is characterized in that: The multi-factor dynamic weight set is specifically: The multi-factor dynamic weight set includes a plurality of weight values, and corresponds to each feature value in the evaluation feature set of confidence data in different fields; wherein the feature value is a score obtained by quantifying the feature in the evaluation feature set; The determination process of the multi-factor dynamic weight set is: A1: Analyze the economic policy stability of major trading countries by collecting and quantifying the adjustment frequency of fiscal and monetary policies and the consistency of policies of trading countries within a preset analysis period; A2: Analyze exchange rate fluctuations and calculate the exchange rate fluctuation range of major trading currencies during the analysis period; A3: Quantify the size of the enterprise and its industry position; A4: Based on the quantification of steps A1 to A4, the multi-factor dynamic weight set is obtained.

8. The international trade enterprise credit evaluation algorithm according to claim 6 is characterized in that: The preliminary credit evaluation results are as follows: The preliminary credit evaluation result is obtained by calculating each feature value in the evaluation feature set and the corresponding initial multi-factor dynamic weight set.

9. The international trade enterprise credit evaluation algorithm according to claim 7 is characterized in that: The obtaining of the updated multi-factor dynamic weight set is specifically: When there are updated feature values ​​of the evaluation feature set of confidence data in different fields, steps A1 to A4 are re-executed to obtain an updated multi-factor dynamic weight set.

10. The international trade enterprise credit evaluation algorithm according to claim 9 is characterized in that: The updating of the credit evaluation result is specifically as follows: Obtaining the updated multi-factor dynamic weight set and each feature value in the updated evaluation feature set of confidence data in different fields; Each feature value in the updated evaluation feature set is calculated with the corresponding updated multi-factor dynamic weight set to obtain an updated credit evaluation result.

Images