Enterprise credit information evaluation system and method

Abstract

The invention discloses an enterprise credit information assessment system and method, and the system comprises a data collection module, a data screening module, a text recognition module and a risk assessment module: the data collection module is used for collecting enterprise public information corresponding to a to-be-assessed enterprise; the data screening module is used for cleaning and screening the enterprise public information according to the consistency among various information in the enterprise public information to obtain to-be-evaluated information; the text recognition module is used for performing text recognition on the to-be-assessed information and extracting multiple pieces of target information corresponding to the to-be-assessed enterprise in the to-be-assessed information; and the risk assessment module is used for analyzing and predicting each piece of target information based on a preset identification model, determining the credit risk of the to-be-assessed enterprise, generating an enterprise credit information report based on the credit risk, and completing credit information assessment of the to-be-assessed enterprise. According to the invention, accurate and efficient evaluation of enterprise credit information is realized.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an enterprise credit information evaluation method, system and storage medium. Background Art

[0002] Corporate credit information is the "credit card" of enterprises in market economic activities. It is not only related to the basic daily operations and market activities of the enterprise itself, but also involves multiple aspects such as enterprise financing, cooperation and government supervision. It is of great significance for enterprises to obtain financial services, participate in market competition, establish business cooperation relationships, etc. Therefore, accurate assessment of corporate credit information is crucial for enterprises.

[0003] At present, the evaluation of corporate credit information usually relies on credit rating agencies to collect information for analysis and evaluation. This evaluation method is not only time-consuming and labor-intensive, but also prone to inaccurate evaluation due to insufficient information collection and insufficient analysis. Therefore, how to accurately and efficiently evaluate corporate credit information is one of the technical issues that need to be solved at present. Summary of the invention

[0004] The main purpose of the present invention is to provide a system and method for evaluating corporate credit information, aiming to solve the technical problem of how to accurately and efficiently evaluate corporate credit information in the prior art.

[0005] To achieve the above-mentioned purpose, the present invention provides an enterprise credit information evaluation system, which includes a data collection module, a data screening module, a text recognition module and a risk evaluation module: wherein,

[0006] The data collection module is used to collect public information of enterprises corresponding to the enterprise to be evaluated, and the public information of enterprises includes but is not limited to basic information of enterprises, news report information, advertising information, and social public opinion information;

[0007] The data screening module is used to clean and screen the enterprise public information according to the consistency between various information in the enterprise public information to obtain the information to be evaluated;

[0008] The text recognition module is used to perform text recognition on the information to be evaluated, and extract multiple target information corresponding to the enterprise to be evaluated from the information to be evaluated;

[0009] The risk assessment module is used to analyze and predict each of the target information based on a preset recognition model, determine the credit risk of the enterprise to be assessed, and generate an enterprise credit information report based on the credit risk to complete the credit information assessment of the enterprise to be assessed.

[0010] Preferably, the enterprise credit information evaluation system further includes:

[0011] A first acquisition module is used to acquire the industry reference weight and the expert reference weight corresponding to each of the target information;

[0012] An updating module is used to update the credit risk according to the industry reference weight and the expert reference weight.

[0013] Preferably, the enterprise credit information evaluation system further includes:

[0014] A second acquisition module is used to acquire an access user account corresponding to the access request when receiving the access request;

[0015] A first judgment module is used to judge whether the account authority corresponding to the access user account matches the report authority corresponding to the enterprise credit information report, and if they match, send the enterprise credit information report to the access user account;

[0016] The second judgment module is used to judge whether the access request carries authorization information when the account authority does not match the report authority. If it carries authorization information, the enterprise credit information report is sent to the access user account; if it does not carry authorization information, a prompt message of denying access is sent to the access user account.

[0017] Preferably, the risk assessment module comprises:

[0018] A judging unit, used to judge whether there is a report template corresponding to the enterprise to be assessed;

[0019] an adding unit, configured to add the credit risk and the associated data corresponding to each field in each of the target information and the information to be evaluated to the report template when the report template exists, according to the correspondence between each field in the report template and the credit risk, so as to generate the enterprise credit information report;

[0020] The adding unit is also used to add the credit risk and the associated data corresponding to the credit risk in each of the target information and the information to be evaluated to the system default template to generate the enterprise credit information report when the report template does not exist.

[0021] Preferably, the enterprise credit information evaluation system further includes:

[0022] A third acquisition module is used to acquire enterprise sample data and divide the enterprise sample data into training sample data and test sample data;

[0023] A training module, used to train a preset initial model based on the training sample data, and when the number of training times reaches a preset number, test the preset initial model based on the test sample data to obtain a test result;

[0024] A calculation module is used to calculate the loss function value of the preset initial model based on the test result, and determine whether the loss function value is less than a preset loss threshold, wherein the calculation formula of the loss function value is:

[0025] L=-[y*log(p)+(1-y)*log(1-p)]

[0026] Among them, y represents the reference result corresponding to the test sample data, and p represents the test result;

[0027] An iterative module is used to generate the preset initial model as a preset recognition model when the loss function value is less than a preset loss threshold, and to execute the step of training the preset initial model based on the training sample data when the loss function value is not less than the preset loss threshold, until the calculated loss function value is less than the preset loss threshold.

[0028] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a method for evaluating corporate credit information, the method comprising:

[0029] Collect the public information of the enterprise corresponding to the enterprise to be evaluated, which includes but is not limited to basic enterprise information, news report information, advertising information, and social public opinion information;

[0030] According to the consistency between various pieces of information in the enterprise public information, the enterprise public information is cleaned and screened to obtain the information to be evaluated;

[0031] Performing text recognition on the information to be evaluated, and extracting a plurality of target information corresponding to the enterprise to be evaluated from the information to be evaluated;

[0032] Based on the preset recognition model, each of the target information is analyzed and predicted to determine the credit risk of the enterprise to be evaluated, and an enterprise credit information report is generated based on the credit risk to complete the credit information evaluation of the enterprise to be evaluated.

[0033] Preferably, the step of generating an enterprise credit information report based on the predicted credit risk includes:

[0034] Obtaining industry reference weights and expert reference weights corresponding to each of the target information;

[0035] The credit risk is updated according to the industry reference weight and the expert reference weight.

[0036] Preferably, the step of generating an enterprise credit information report based on the predicted credit risk comprises:

[0037] When receiving an access request, obtaining an access user account corresponding to the access request;

[0038] Determine whether the account authority corresponding to the access user account matches the report authority corresponding to the enterprise credit information report, and if so, send the enterprise credit information report to the access user account;

[0039] If the account authority does not match the report authority, determine whether the access request carries authorization information. If it carries authorization information, send the enterprise credit information report to the access user account. If it does not carry authorization information, send a prompt message of access denial to the access user account.

[0040] Preferably, the step of generating an enterprise credit information report based on the predicted credit risk includes:

[0041] Determine whether there is a report template corresponding to the enterprise to be assessed;

[0042] If the report template exists, then according to the correspondence between each field in the report template and the credit risk, the credit risk and the associated data corresponding to each field in each of the target information and the information to be evaluated are added to the report template to generate the enterprise credit information report;

[0043] If the report template does not exist, the credit risk and the associated data corresponding to the credit risk in each of the target information and the information to be evaluated are added to the system default template to generate the enterprise credit information report.

[0044] Preferably, before the step of analyzing and predicting each of the target information based on a preset recognition model to determine the credit risk of the enterprise to be assessed, the step includes:

[0045] Acquire enterprise sample data, and divide the enterprise sample data into training sample data and test sample data;

[0046] The preset initial model is trained based on the training sample data, and when the number of training times reaches a preset number, the preset initial model is tested based on the test sample data to obtain a test result;

[0047] The loss function value of the preset initial model is calculated based on the test result, and it is determined whether the loss function value is less than a preset loss threshold, wherein the calculation formula of the loss function value is:

[0048] L=-[y*log(p)+(1-y)*log(1-p)]

[0049] Among them, y represents the reference result corresponding to the test sample data, and p represents the test result;

[0050] If the loss function value is less than a preset loss threshold, generating the preset initial model as a preset recognition model;

[0051] If the loss function value is not less than the preset loss threshold, the step of training the preset initial model based on the training sample data is executed until the calculated loss function value is less than the preset loss threshold.

[0052] The enterprise credit information evaluation system and method of the present invention comprises a data collection module, a data screening module, a text recognition module and a risk assessment module: wherein the data collection module is used to collect enterprise public information corresponding to the enterprise to be evaluated, and the collected enterprise public data includes but is not limited to enterprise basic information, news report information, advertising information, and social public opinion information; the data screening module is used to clean and screen the enterprise public information according to the consistency between various information in the enterprise public information, and obtain the information to be evaluated; the text recognition module is used to perform text recognition on the information to be evaluated, and extract multiple target information corresponding to the enterprise to be evaluated in the information to be evaluated; the risk assessment module is used to analyze and predict each target information based on a preset recognition model, determine the credit risk of the enterprise to be evaluated, and generate an enterprise credit information report based on the predicted credit risk, and complete the credit information evaluation of the enterprise to be evaluated. In this way, various types of public information related to the enterprise are collected by means of crawler technology, so that the collected information is more efficient and comprehensive, and the target information obtained after cleaning, screening and identifying various types of public information is predicted and evaluated by a preset recognition model formed by a large amount of data training, so that the generated enterprise credit information report is faster and more accurate, and an accurate and efficient evaluation of the enterprise credit information is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 This is a module diagram of the first embodiment of the enterprise credit information evaluation system of the present invention;

[0054] Figure 2 This is a module diagram of the second embodiment of the enterprise credit information evaluation system of the present invention;

[0055] Figure 3 This is a flow chart of the first embodiment of the enterprise credit information evaluation method of the present invention;

[0056] Figure 4 This is a flow chart of the second embodiment of the enterprise credit information evaluation method of the present invention.

[0057] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0058] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0059] The present invention provides a corporate credit information evaluation system, please refer to Figure 1 , Figure 1 This is a module diagram of the first embodiment of the enterprise credit information evaluation system of the present invention.

[0060] Specifically, the enterprise credit information evaluation system in this embodiment includes a data collection module 10, a data screening module 20, a text recognition module 30 and a risk evaluation module 40, wherein:

[0061] The data collection module 10 is used to collect public information of enterprises corresponding to the enterprise to be evaluated, and the public information of enterprises includes but is not limited to basic information of enterprises, news report information, advertising information, and social public opinion information;

[0062] Furthermore, when an enterprise needs to conduct a credit information evaluation, it initiates an evaluation request to the enterprise credit information evaluation system of this embodiment. After receiving the evaluation request, the enterprise credit information evaluation system reads the enterprise ID (Identification) carried therein, determines the specific enterprise that needs to undergo credit information evaluation by the enterprise ID, and takes the enterprise that needs to undergo credit information evaluation as the enterprise to be evaluated.

[0063] Furthermore, the enterprise credit information evaluation system initiates the enterprise credit information evaluation for the enterprise to be evaluated, and the data collection module therein collects various information related to the enterprise to be evaluated from various official websites, news websites, social platforms and other channels as enterprise public information by calling API (Application Programming Interface), web crawlers and other technologies. Among them, according to the nature of the website from which the various information comes and the category of the information itself, the collected enterprise public information at least includes basic enterprise information indicating the nature and scale of the enterprise, news report information, advertising information and social public opinion information.

[0064] The data screening module 20 is used to clean and screen the enterprise public information according to the consistency between various information in the enterprise public information to obtain the information to be evaluated;

[0065] Understandably, the accuracy of various information disclosed on the Internet varies. Therefore, for the enterprise public information collected by the collection module, the data screening module cleans and screens it according to the consistency between the various information in the enterprise public information. Consistency is used to indicate whether there is a contradiction between the same type of data from different websites. If there is no contradiction, it means that the consistency between the data is high and the data is relatively accurate. If there is a contradiction, it means that the consistency between the data is low, and accurate data needs to be screened out. The order of screening is official data first, that is, official data shall prevail. After cleaning the data irrelevant to the enterprise, removing duplicates, and screening out accurate data, the information to be evaluated related to the enterprise is obtained.

[0066] The text recognition module 30 is used to perform text recognition on the information to be evaluated, and extract multiple target information corresponding to the enterprise to be evaluated from the information to be evaluated;

[0067] Furthermore, in addition to information related to the enterprise's credit information, the information to be evaluated may also contain other information about the enterprise. In order to accurately evaluate the enterprise's credit information, the text recognition module performs text recognition on the information to be evaluated, and extracts information related to the enterprise's credit information from the information to be evaluated through text recognition as multiple target information corresponding to the enterprise to be evaluated.

[0068] The risk assessment module 40 is used to analyze and predict each of the target information based on a preset recognition model, determine the credit risk of the enterprise to be assessed, and generate an enterprise credit information report based on the credit risk to complete the credit information assessment of the enterprise to be assessed.

[0069] Furthermore, the enterprise credit information evaluation system has a preset recognition model formed in advance through training. The risk assessment module uses the preset recognition model to analyze and predict each target information to determine the credit risk of the enterprise to be evaluated. In addition, the determined credit risk and the determined related basis are jointly generated into an enterprise credit information report to complete the credit information evaluation of the enterprise to be evaluated.

[0070] The enterprise credit information report can be generated based on a template pre-set in the enterprise credit information evaluation system, or can be generated by the enterprise to be evaluated selecting a template style. Specifically, the risk assessment module includes:

[0071] A determination unit 41 is used to determine whether there is a report template corresponding to the enterprise to be evaluated;

[0072] An adding unit 42 is used for adding the credit risk and the associated data corresponding to each field in each of the target information and the information to be evaluated to the report template when the report template exists, according to the correspondence between each field in the report template and the credit risk, to generate the enterprise credit information report;

[0073] The adding module 42 is also used to add the credit risk and the associated data corresponding to the credit risk in each of the target information and the information to be evaluated to the system default template to generate the enterprise credit information report when the report template does not exist.

[0074] Furthermore, when initiating an evaluation request, the enterprise to be evaluated can choose whether to generate an enterprise credit information report in accordance with the style specified by it. If the specified style is selected, the specified style is added to the evaluation request. The enterprise credit information evaluation system stores the specified style as a report template corresponding to the enterprise to be evaluated. When the risk evaluation module generates an enterprise credit information report based on credit risk, it first determines whether a report template corresponding to the enterprise to be evaluated is stored through the judgment unit therein. If the report template is stored, the credit risk, as well as the associated data corresponding to each field in each target information and the information to be evaluated are added to the report template through the adding unit according to the correspondence between each field in the report template and the credit risk, to generate the final enterprise credit information report. Among them, the target information is added to the report template as one of the bases for determining the credit risk, and the fields in the report template reflect the content items that the enterprise to be evaluated is concerned about, and the associated data corresponding to it in the information to be evaluated also reflect the basis for generating credit risk that the enterprise to be evaluated is concerned about. The enterprise credit information report thus generated not only accurately evaluates the credit information of the enterprise, but also meets the user's concerns and is convenient for the user to view.

[0075] Furthermore, if it is determined that there is no report template stored in the enterprise credit information evaluation system, the system default template is used to generate the report. The credit risk and the associated data corresponding to the credit risk in each target information and the information to be evaluated are added to the system default template through the adding unit to generate the final enterprise credit information report.

[0076] It should be noted that the credit risk generated by analyzing and predicting the preset identification model may be inaccurate, which makes the enterprise credit information report generated based on the credit risk inaccurate. Therefore, in order to ensure the accuracy of the credit risk, this embodiment also modifies and updates the generated credit risk based on the industry in which the enterprise to be evaluated is located and the expert experience in the industry. Specifically, the enterprise credit information evaluation system also includes:

[0077] The first acquisition module a1 is used to acquire the industry reference weight and the expert reference weight corresponding to each of the target information;

[0078] The updating module a2 is used to update the credit risk according to the industry reference weight and the expert reference weight.

[0079] Furthermore, industry reference weights and expert reference weights are set in advance for each type of target information in each industry based on historical data. After analyzing and predicting the target information of the enterprise to be evaluated based on the preset recognition model and determining the credit risk of the enterprise to be evaluated, the industry reference weights and expert reference weights corresponding to each target information are acquired through the first acquisition module. Then, based on the acquired industry reference weights and expert reference weights, the determined credit risk is updated through the update module to make the determined credit risk more accurate, thereby ensuring the accuracy of the generated enterprise credit information report.

[0080] It is understandable that the enterprise to be evaluated has a need to view the generated enterprise credit information report. In order to ensure the security of viewing, this embodiment also provides a viewing permission mechanism. Specifically, the enterprise credit information evaluation system also includes:

[0081] The second acquisition module b1 is used to acquire the access user account corresponding to the access request when receiving the access request;

[0082] A first judgment module b2 is used to judge whether the account authority corresponding to the access user account matches the report authority corresponding to the enterprise credit information report, and if they match, send the enterprise credit information report to the access user account;

[0083] The second judgment module b3 is used to determine whether the access request carries authorization information when the account authority does not match the report authority. If it carries authorization information, the enterprise credit information report is sent to the access user account. If it does not carry authorization information, a prompt message of denying access is sent to the access user account.

[0084] Further, when the enterprise credit information system receives an access request for the enterprise credit information report of the enterprise to be evaluated, the access user account corresponding to the access request is obtained through the second acquisition module. Then, the first judgment module is used to determine whether the account authority corresponding to the access user account matches the report authority corresponding to the enterprise credit information report. If they match, it means that the access user account can access the enterprise credit information report, so the enterprise credit information report is sent to the access user account. On the contrary, if it is determined that the two permissions do not match, it means that the access user account does not have the authority to access the enterprise credit information report. At this time, it is necessary to determine whether the access request initiated by the access user account carries authorization information through the second judgment module. If it carries authorization information, it means that although the access user account itself does not have the authority to access the enterprise credit information report, it is authorized by the account with access rights, so that the access user account has access weight. Therefore, the enterprise credit information report is sent to the access user account. If it is determined that the access request does not carry authorization information, it means that the access user account itself does not have access rights and is not authorized, so a prompt message of denying access is sent to the access user account. The security of enterprise credit information is ensured through permission setting.

[0085] The enterprise credit information evaluation system implemented in this embodiment includes a data collection module, a data screening module, a text recognition module and a risk assessment module: wherein the data collection module is used to collect the enterprise public information corresponding to the enterprise to be evaluated, and the collected enterprise public data includes but is not limited to enterprise basic information, news report information, advertising information, and social public opinion information; the data screening module is used to clean and screen the enterprise public information according to the consistency between the various information in the enterprise public information, and obtain the information to be evaluated; the text recognition module is used to perform text recognition on the information to be evaluated, and extract multiple target information corresponding to the enterprise to be evaluated in the information to be evaluated; the risk assessment module is used to analyze and predict each target information based on the preset recognition model, determine the credit risk of the enterprise to be evaluated, and generate an enterprise credit information report based on the predicted credit risk, and complete the credit information evaluation of the enterprise to be evaluated. In this way, various types of public information related to the enterprise are collected with the help of crawler technology, making the collected information more efficient and comprehensive, and the target information obtained after cleaning, screening and identifying various types of public information is predicted and evaluated by the preset recognition model formed by a large amount of data training, so that the generated enterprise credit information report is faster and more accurate, and the accurate and efficient evaluation of enterprise credit information is achieved.

[0086] For further information, please refer to Figure 2 Based on the first embodiment of the enterprise credit information evaluation system of the present invention, a second embodiment of the enterprise credit information evaluation system of the present invention is proposed.

[0087] The difference between the second embodiment of the enterprise credit information evaluation system and the first embodiment of the enterprise credit information evaluation system is that the enterprise credit information evaluation system further includes:

[0088] A third acquisition module 50 is used to acquire enterprise sample data and divide the enterprise sample data into training sample data and test sample data;

[0089] A training module 60 is used to train a preset initial model based on the training sample data, and when the number of training times reaches a preset number, test the preset initial model based on the test sample data to obtain a test result;

[0090] A calculation module 70, configured to calculate a loss function value of the preset initial model based on the test result, and determine whether the loss function value is less than a preset loss threshold;

[0091] The iteration module 80 is used to generate the preset initial model as a preset recognition model when the loss function value is less than a preset loss threshold, and to execute the step of training the preset initial model based on the training sample data when the loss function value is not less than the preset loss threshold, until the calculated loss function value is less than the preset loss threshold.

[0092] This embodiment generates a preset recognition model by training with a large amount of enterprise sample data. Specifically, a large amount of enterprise sample data is first obtained through the third acquisition module, and the enterprise sample data is divided into training sample data and test sample data according to a preset ratio, for example, according to a ratio of 7:3 or 8:2. Then, the training module trains the preset initial model based on the training sample data, and counts the number of trainings. When the counted number of trainings reaches a preset number, the preset initial model is tested with the test sample data, and the test sample data is processed by the preset initial model to obtain the processed test results. Then, the calculation module calculates the loss function value of the preset initial model based on the test results. The calculation formula can be seen in the following formula (1):

[0093] L=-[y*log(p)+(1-y)*log(1-p)]

[0094] Among them, y represents the reference result corresponding to the test sample data, and p represents the test result.

[0095] Furthermore, in order to reflect the loss size of the loss function value, a preset loss threshold is pre-set, and the loss function value is compared with the preset loss threshold to determine whether the loss function value is less than the preset loss threshold. If it is less than the preset loss threshold, it means that the error between the processing result obtained by the preset initial model for processing the test sample data and the reference result is small, and the preset initial model has better processing performance, so the preset initial model is generated as the preset recognition model. On the contrary, if it is determined through comparison that the loss function value is greater than or equal to the preset loss threshold, it means that the processing performance of the preset initial model is poor. At this time, the iterative training of the preset initial model is controlled by the iteration module, and the steps of training the preset initial model according to the training sample data are executed, and the test sample data is re-tested and a new loss function value is calculated until the calculated loss function value is less than the preset loss threshold, the iterative training of the preset initial model is stopped, and the preset initial model is generated as the preset recognition model.

[0096] This embodiment generates a preset recognition model through training with a large amount of enterprise sample data, so that the preset recognition model can accurately analyze the target information, improve the accuracy of the determined enterprise credit risk, and thus make the enterprise's credit information assessment more accurate.

[0097] In addition, the present invention provides a method for evaluating corporate credit information. Figure 3 , Figure 3 It is a flowchart of the first embodiment of the enterprise credit information evaluation method of the present invention. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described can be performed in a different order than here.

[0098] Specifically, the enterprise credit information evaluation method in this embodiment includes:

[0099] Step S10, collecting public information of the enterprise corresponding to the enterprise to be evaluated, wherein the public information of the enterprise includes but is not limited to basic information of the enterprise, news report information, advertising information, and social public opinion information;

[0100] The enterprise credit information evaluation method of this embodiment is applied to the enterprise credit information evaluation system. When an enterprise needs to conduct a credit information evaluation, it initiates an evaluation request to the enterprise credit information evaluation system. After receiving the evaluation request, the enterprise credit information evaluation system reads the enterprise ID (Identification) carried therein, determines the specific enterprise that needs to conduct a credit information evaluation based on the enterprise ID, and takes the enterprise that needs to conduct a credit information evaluation as the enterprise to be evaluated.

[0101] Furthermore, the enterprise credit information evaluation system starts the enterprise credit information evaluation for the enterprise to be evaluated, and first collects various information related to the enterprise to be evaluated from various official websites, news websites, social platforms and other channels as enterprise public information by calling API, web crawlers and other technologies. Among them, according to the nature of the website where the various information comes from and the category of the information itself, the collected enterprise public information at least includes basic enterprise information indicating the nature and scale of the enterprise, news report information, advertising information and social public opinion information.

[0102] Step S20, cleaning and screening the enterprise public information according to the consistency between various information in the enterprise public information to obtain information to be evaluated;

[0103] Understandably, the accuracy of various information disclosed on the Internet varies. Therefore, for the collected public information of enterprises, it can be cleaned and screened based on the consistency between the various information in the public information of enterprises. Consistency is used to indicate whether there is a contradiction between the same type of data from different websites. If there is no contradiction, it means that the consistency between the data is high and the data is relatively accurate. If there is a contradiction, it means that the consistency between the data is low, and accurate data needs to be screened out. The order of screening is official data first, that is, official data shall prevail. After cleaning the data irrelevant to the enterprise, removing duplicates, and screening out accurate data, the information to be evaluated related to the enterprise is obtained.

[0104] Step S30, performing text recognition on the information to be evaluated, and extracting a plurality of target information corresponding to the enterprise to be evaluated from the information to be evaluated;

[0105] Furthermore, the information to be evaluated may contain other information of the enterprise in addition to information related to the enterprise's credit information. In order to accurately evaluate the enterprise's credit information, text recognition is performed on the information to be evaluated, and information related to the enterprise's credit information in the information to be evaluated is extracted through text recognition as multiple target information corresponding to the enterprise to be evaluated.

[0106] Step S40, analyzing and predicting each of the target information based on a preset recognition model, determining the credit risk of the enterprise to be evaluated, and generating an enterprise credit information report based on the credit risk, thereby completing the credit information evaluation of the enterprise to be evaluated.

[0107] Furthermore, the enterprise credit information evaluation system has a preset recognition model formed in advance through training, and each target information is analyzed and predicted through the preset recognition model to determine the credit risk of the enterprise to be evaluated. In addition, the determined credit risk and the related basis are jointly generated into an enterprise credit information report to complete the credit information evaluation of the enterprise to be evaluated.

[0108] The enterprise credit information report can be generated based on a template pre-set in the enterprise credit information evaluation system, or can be generated by the enterprise to be evaluated selecting a template style. Specifically, the steps of generating an enterprise credit information report based on predicted credit risk include:

[0109] Step S41, determining whether there is a report template corresponding to the enterprise to be evaluated;

[0110] Step S42, if the report template exists, then according to the correspondence between each field in the report template and the credit risk, the credit risk and the associated data corresponding to each field in each of the target information and the information to be evaluated are added to the report template to generate the enterprise credit information report;

[0111] Step S43: If the report template does not exist, the credit risk and the associated data corresponding to the credit risk in each of the target information and the information to be evaluated are added to the system default template to generate the enterprise credit information report.

[0112] Furthermore, when initiating an evaluation request, the enterprise to be evaluated can choose whether to generate an enterprise credit information report in accordance with the style specified by it. If the specified style is selected, the specified style is added to the evaluation request. The enterprise credit information evaluation system stores the specified style as a report template corresponding to the enterprise to be evaluated. When generating an enterprise credit information report based on credit risk, it is first determined whether a report template corresponding to the enterprise to be evaluated is stored. If the report template is stored, the credit risk, as well as the associated data corresponding to each field in each target information and the information to be evaluated are added to the report template according to the correspondence between each field in the report template and the credit risk, to generate the final enterprise credit information report. Among them, the target information is added to the report template as one of the bases for determining the credit risk, and the fields in the report template reflect the content items that the enterprise to be evaluated is concerned about, and the associated data corresponding to it in the information to be evaluated also reflect the basis for generating credit risk that the enterprise to be evaluated is concerned about. The enterprise credit information report thus generated not only accurately evaluates the credit information of the enterprise, but also meets the user's concerns and is convenient for users to view.

[0113] Furthermore, if it is determined that there is no report template stored in the enterprise credit information evaluation system, the system default template is used to generate the report. The credit risk and the associated data corresponding to the credit risk in each target information and the information to be evaluated are added to the system default template to generate the final enterprise credit information report.

[0114] It should be noted that the credit risk generated by analyzing and predicting the preset identification model may be inaccurate, which makes the corporate credit information report generated based on the credit risk inaccurate. Therefore, in order to ensure the accuracy of the credit risk, this embodiment also modifies and updates the generated credit risk based on the industry in which the enterprise to be evaluated is located and the expert experience in the industry. Specifically, the step of generating the corporate credit information report based on the predicted credit risk includes:

[0115] Step c1, obtaining the industry reference weight and expert reference weight corresponding to each of the target information;

[0116] Step c2: updating the credit risk according to the industry reference weight and the expert reference weight.

[0117] Furthermore, industry reference weights and expert reference weights are set in advance for each type of target information in each industry based on historical data. After analyzing and predicting the target information of the enterprise to be evaluated and determining the credit risk of the enterprise to be evaluated, the industry reference weights and expert reference weights corresponding to each target information are obtained. Then, based on the obtained industry reference weights and expert reference weights, the determined credit risk is updated to make the determined credit risk more accurate, thereby ensuring the accuracy of the generated enterprise credit information report.

[0118] It is understandable that the enterprise to be assessed has a need to view the generated enterprise credit information report. In order to ensure the security of viewing, this embodiment also provides a viewing permission mechanism. Specifically, the step of generating an enterprise credit information report based on the predicted credit risk includes:

[0119] Step d1, when an access request is received, obtaining an access user account corresponding to the access request;

[0120] Step d2, determining whether the account authority corresponding to the access user account matches the report authority corresponding to the enterprise credit information report, and if so, sending the enterprise credit information report to the access user account;

[0121] Step d3, if the account authority does not match the report authority, determine whether the access request carries authorization information. If it carries authorization information, send the enterprise credit information report to the access user account. If it does not carry authorization information, send a prompt message of access denial to the access user account.

[0122] Further, when the enterprise credit information system receives an access request for the enterprise credit information report of the enterprise to be evaluated, the access user account corresponding to the access request is obtained. Then it is determined whether the account authority corresponding to the access user account matches the report authority corresponding to the enterprise credit information report. If they match, it means that the access user account can access the enterprise credit information report, so the enterprise credit information report is sent to the access user account. On the contrary, if it is determined that the two permissions do not match, it means that the access user account does not have the authority to access the enterprise credit information report. At this time, it is necessary to determine whether the access request initiated by the access user account carries authorization information. If it carries authorization information, it means that although the access user account itself does not have the authority to access the enterprise credit information report, it is authorized by the account with access rights, so that the access user account has access weight. Therefore, the enterprise credit information report is sent to the access user account. If it is determined that the access request does not carry authorization information, it means that the access user account itself does not have access rights and is not authorized, so a prompt message of access denial is sent to the access user account. The security of enterprise credit information is ensured through permission setting.

[0123] The enterprise credit information evaluation method implemented in this implementation first collects the enterprise public information corresponding to the enterprise to be evaluated. The collected enterprise public data includes but is not limited to enterprise basic information, news report information, advertising information, and social public opinion information; then, according to the consistency between the various information in the enterprise public information, the enterprise public information is cleaned and screened to obtain the information to be evaluated; then, the information to be evaluated is subjected to text recognition to extract multiple target information corresponding to the enterprise to be evaluated in the information to be evaluated; and each target information is analyzed and predicted based on the preset recognition model to determine the credit risk of the enterprise to be evaluated, and the enterprise credit information report is generated based on the predicted credit risk to complete the credit information evaluation of the enterprise to be evaluated. In this way, various types of public information related to the enterprise are collected with the help of crawler technology, making the collected information more efficient and comprehensive, and the target information obtained after cleaning, screening and identifying various types of public information is predicted and evaluated by the preset recognition model formed by a large amount of data training, so that the generated enterprise credit information report is faster and more accurate, and the accurate and efficient evaluation of enterprise credit information is achieved.

[0124] For further information, please refer to Figure 4 Based on the first embodiment of the enterprise credit information evaluation method of the present invention, a second embodiment of the enterprise credit information evaluation method of the present invention is proposed.

[0125] The difference between the second embodiment of the enterprise credit information evaluation method and the first embodiment of the enterprise credit information evaluation method is that before the step of analyzing and predicting each of the target information based on the preset recognition model to determine the credit risk of the enterprise to be evaluated includes:

[0126] Step S50, obtaining enterprise sample data, and dividing the enterprise sample data into training sample data and test sample data;

[0127] Step S50, training a preset initial model based on the training sample data, and when the number of training times reaches a preset number, testing the preset initial model based on the test sample data to obtain a test result;

[0128] Step S60, calculating the loss function value of the preset initial model based on the test result, and determining whether the loss function value is less than a preset loss threshold;

[0129] Step S70, if the loss function value is less than a preset loss threshold, generating the preset initial model as a preset recognition model;

[0130] Step S80, if the loss function value is not less than the preset loss threshold, execute the step of training the preset initial model based on the training sample data until the calculated loss function value is less than the preset loss threshold.

[0131] This embodiment generates a preset recognition model by training with a large amount of enterprise sample data. Specifically, a large amount of enterprise sample data is first obtained, and the enterprise sample data is divided into training sample data and test sample data according to a preset ratio, for example, divided according to a ratio of 7:3 or 8:2. Then, the preset initial model is trained based on the training sample data, and the number of trainings is counted. When the number of trainings counted reaches a preset number, the preset initial model is tested with the test sample data, and the test sample data is processed by the preset initial model to obtain the processed test results. Then, the loss function value of the preset initial model is calculated based on the test results. The calculation formula can be found in the above formula (1), which will not be repeated here.

[0132] Furthermore, in order to reflect the loss size of the loss function value, a preset loss threshold is pre-set, and the loss function value is compared with the preset loss threshold to determine whether the loss function value is less than the preset loss threshold. If it is less than the preset loss threshold, it means that the error between the processing result obtained by the preset initial model for processing the test sample data and the reference result is small, and the preset initial model has good processing performance, so the preset initial model is generated as the preset recognition model. On the contrary, if it is determined through comparison that the loss function value is greater than or equal to the preset loss threshold, it means that the processing performance of the preset initial model is poor. At this time, the preset initial model is iteratively trained, and the steps of training the preset initial model according to the training sample data are executed, and the test sample data is re-tested and a new loss function value is calculated until the calculated loss function value is less than the preset loss threshold, the iterative training of the preset initial model is stopped, and the preset initial model is generated as the preset recognition model.

[0133] This embodiment generates a preset recognition model through training with a large amount of enterprise sample data, so that the preset recognition model can accurately analyze the target information, improve the accuracy of the determined enterprise credit risk, and thus make the enterprise's credit information assessment more accurate.

[0134] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims. All equivalent structures or equivalent process changes made using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are protected by the present invention.

Claims

1. An enterprise credit information evaluation system, characterized in that: It includes data collection module, data screening module, text recognition module and risk assessment module: among which, The data collection module is used to collect public information of enterprises corresponding to the enterprise to be evaluated, and the public information of enterprises includes but is not limited to basic information of enterprises, news report information, advertising information, and social public opinion information; The data screening module is used to clean and screen the enterprise public information according to the consistency between various information in the enterprise public information to obtain the information to be evaluated; The text recognition module is used to perform text recognition on the information to be evaluated, and extract multiple target information corresponding to the enterprise to be evaluated from the information to be evaluated; The risk assessment module is used to analyze and predict each of the target information based on a preset recognition model, determine the credit risk of the enterprise to be assessed, and generate an enterprise credit information report based on the credit risk to complete the credit information assessment of the enterprise to be assessed.

2. The enterprise credit information evaluation system according to claim 1, characterized in that: The enterprise credit information evaluation system also includes: A first acquisition module is used to acquire the industry reference weight and the expert reference weight corresponding to each of the target information; An updating module is used to update the credit risk according to the industry reference weight and the expert reference weight.

3. The enterprise credit information evaluation system according to claim 1, characterized in that: The enterprise credit information evaluation system also includes: A second acquisition module is used to acquire an access user account corresponding to the access request when receiving the access request; A first judgment module is used to judge whether the account authority corresponding to the access user account matches the report authority corresponding to the enterprise credit information report, and if they match, send the enterprise credit information report to the access user account; The second judgment module is used to judge whether the access request carries authorization information when the account authority does not match the report authority. If it carries authorization information, the enterprise credit information report is sent to the access user account; if it does not carry authorization information, a prompt message of denying access is sent to the access user account.

4. The enterprise credit information evaluation system according to any one of claims 1 to 3, characterized in that: The risk assessment module includes: A judging unit, used to judge whether there is a report template corresponding to the enterprise to be assessed; an adding unit, configured to add the credit risk and the associated data corresponding to each field in each of the target information and the information to be evaluated to the report template when the report template exists, according to the correspondence between each field in the report template and the credit risk, to generate the enterprise credit information report; The adding unit is also used to add the credit risk and the associated data corresponding to the credit risk in each of the target information and the information to be evaluated to the system default template to generate the enterprise credit information report when the report template does not exist.

5. The enterprise credit information evaluation system according to any one of claims 1 to 3, characterized in that: The enterprise credit information evaluation system also includes: A third acquisition module is used to acquire enterprise sample data and divide the enterprise sample data into training sample data and test sample data; A training module is used to train a preset initial model based on the training sample data, and when the number of training times reaches a preset number, test the preset initial model based on the test sample data to obtain a test result; A calculation module is used to calculate the loss function value of the preset initial model based on the test result, and determine whether the loss function value is less than a preset loss threshold, wherein the calculation formula of the loss function value is: L=-[y*log(p)+(1-y)*log(1-p)] Among them, y represents the reference result corresponding to the test sample data, and p represents the test result; An iterative module is used to generate the preset initial model as a preset recognition model when the loss function value is less than a preset loss threshold, and to execute the step of training the preset initial model based on the training sample data when the loss function value is not less than the preset loss threshold, until the calculated loss function value is less than the preset loss threshold.

6. A method for evaluating corporate credit information, characterized in that: The enterprise credit information evaluation method comprises: Collect the public information of the enterprise corresponding to the enterprise to be evaluated, which includes but is not limited to basic enterprise information, news report information, advertising information, and social public opinion information; According to the consistency between various pieces of information in the enterprise public information, the enterprise public information is cleaned and screened to obtain the information to be evaluated; Performing text recognition on the information to be evaluated, and extracting a plurality of target information corresponding to the enterprise to be evaluated from the information to be evaluated; Based on the preset recognition model, each of the target information is analyzed and predicted to determine the credit risk of the enterprise to be evaluated, and an enterprise credit information report is generated based on the credit risk to complete the credit information evaluation of the enterprise to be evaluated.

7. The enterprise credit information evaluation method according to claim 6, characterized in that: The step of generating an enterprise credit information report based on the predicted credit risk includes: Obtaining industry reference weights and expert reference weights corresponding to each of the target information; The credit risk is updated according to the industry reference weight and the expert reference weight.

8. The enterprise credit information evaluation method according to claim 6, characterized in that: The step of generating an enterprise credit information report based on the predicted credit risk then includes: When receiving an access request, obtaining an access user account corresponding to the access request; Determine whether the account authority corresponding to the access user account matches the report authority corresponding to the enterprise credit information report, and if so, send the enterprise credit information report to the access user account; If the account authority does not match the report authority, determine whether the access request carries authorization information. If it carries authorization information, send the enterprise credit information report to the access user account. If it does not carry authorization information, send a prompt message of access denial to the access user account.

9. The enterprise credit information evaluation method according to any one of claims 6 to 8, characterized in that: The step of generating an enterprise credit information report based on the predicted credit risk includes: Determine whether there is a report template corresponding to the enterprise to be assessed; If the report template exists, then according to the correspondence between each field in the report template and the credit risk, the credit risk and the associated data corresponding to each field in each of the target information and the information to be evaluated are added to the report template to generate the enterprise credit information report; If the report template does not exist, the credit risk and the associated data corresponding to the credit risk in each of the target information and the information to be evaluated are added to the system default template to generate the enterprise credit information report.

10. The enterprise credit information evaluation method according to any one of claims 6 to 8, characterized in that: The step of analyzing and predicting each of the target information based on the preset recognition model to determine the credit risk of the enterprise to be assessed includes: Acquire enterprise sample data, and divide the enterprise sample data into training sample data and test sample data; The preset initial model is trained based on the training sample data, and when the number of training times reaches a preset number, the preset initial model is tested based on the test sample data to obtain a test result; The loss function value of the preset initial model is calculated based on the test result, and it is determined whether the loss function value is less than a preset loss threshold, wherein the calculation formula of the loss function value is: L=-[y*log(p)+(1-y)*log(1-p)] Among them, y represents the reference result corresponding to the test sample data, and p represents the test result; If the loss function value is less than a preset loss threshold, generating the preset initial model as a preset recognition model; If the loss function value is not less than the preset loss threshold, the step of training the preset initial model based on the training sample data is executed until the calculated loss function value is less than the preset loss threshold.

Images