An enterprise credit risk prediction system and method based on big data

Abstract

The present invention discloses a big data-based enterprise credit risk prediction system and method. The big data-based enterprise credit risk prediction system is provided with an MPP data service layer, wherein the MPP data service layer is provided with an enterprise project comprehensive evaluation module, an enterprise credit comprehensive evaluation module, and an enterprise project credit evaluation module; the enterprise project comprehensive evaluation module outputs an enterprise project comprehensive evaluation result x; the enterprise credit comprehensive evaluation module outputs an enterprise credit comprehensive evaluation result z; and the enterprise project credit evaluation module outputs an enterprise project credit evaluation result y based on the enterprise project comprehensive evaluation result x and the enterprise credit comprehensive evaluation result z. Beneficial effects: The present invention adds project evaluation, which is mainly based on historical project data and supplemented by new project and current project information. It predicts the feasibility and future development status of the client enterprise's new project, and then combines the client enterprise's credit evaluation results to improve the accuracy and reliability of the evaluation results and reduce the working risks of financial institutions.

Description

Technical Field

[0001] The present invention relates to the field of financial risk technology, and in particular to a system and method for predicting corporate credit risk based on big data. Background Art

[0002] In today's digital age, the rapid development of big data technology has profoundly transformed the financial industry's operations and business models. Big data not only presents unprecedented opportunities across various industries, but also raises a host of new challenges. Within the financial sector, corporate credit and investment, as crucial components of financial services, are also deeply impacted by big data. The scale of credit assets and the number of clients of financial institutions have climbed year by year, and the number of corporate and factory credit clients has doubled. The application of big data technology has become a key tool for improving efficiency and reducing risk.

[0003] At present, financial institutions use big data to obtain basic information such as registration costs, judicial information, business information, and business risks of client companies, and convert traditional paper information into digital information. At the same time, the evaluation model combines the above digital information to evaluate client companies, and determines the client companies' liquidity and repayment capabilities based on the evaluation results, which greatly reduces the financial institutions' work in collecting client company information and credit assessment, thereby improving the work efficiency of financial institutions.

[0004] However, the existing corporate credit assessment methods have some shortcomings. They are limited to collecting historical information such as the client company's registration costs, judicial information, operating information, and operating risks, and conducting credit assessments on the client company based on the above historical information. This leads to one-sided assessment results and makes it difficult to fully understand the client company's current true credit status.

[0005] At the same time, credit assessment can only reflect the historical credit quality of the client company, and cannot predict the future development and credit status of the client company, causing financial institutions to face a series of problems in risk management. Summary of the Invention

[0006] The purpose of the present invention is to provide a corporate credit risk prediction system and method based on big data, which can predict the feasibility and future development of new projects of future client enterprises, improve the accuracy and reliability of corporate customer credit assessment results, and thus reduce the working risks of financial institutions.

[0007] A big data-based enterprise credit risk prediction system includes an original data layer, a data standard layer, and a data fusion layer. The data standard layer obtains original data from the original data layer, processes the original data, obtains standard data, and sends the standard data to the data fusion layer. The data fusion layer integrates and classifies all standard data, dividing the standard data into at least one category. The key point is that it also includes an MPP data service layer, which is equipped with an enterprise project comprehensive evaluation module, an enterprise credit comprehensive evaluation module, and an enterprise project credit evaluation module.

[0008] The enterprise project comprehensive evaluation module is provided with an enterprise project comprehensive evaluation digital model, which outputs the enterprise project comprehensive evaluation result x;

[0009] The enterprise credit comprehensive evaluation module is provided with an enterprise credit comprehensive evaluation digital model, which outputs the enterprise credit comprehensive evaluation result z;

[0010] The enterprise project credit evaluation module is provided with an enterprise project credit evaluation digital model, which outputs an enterprise project credit evaluation result y according to the enterprise project comprehensive evaluation result x and the enterprise credit comprehensive evaluation result z.

[0011] Through the above structure, a big data-based corporate credit risk prediction system can infer the feasibility and future development of a new project to be launched by a client company based on the client company's historical projects and corporate credit status, thereby improving the accuracy of the company's project credit assessment.

[0012] Furthermore, the original data layer includes a data source, which sends the original data to a buffer layer, and the buffer layer then transmits the original data to the data standard layer.

[0013] Furthermore, a data processing module is provided in the data standard layer, and the data processing module includes a pre-processing unit and a quality inspection unit;

[0014] The preprocessing unit obtains the original data and performs space removal and special value removal processing on the original data to obtain preprocessed data;

[0015] The quality inspection unit is connected to a manual intervention unit, and the quality inspection unit inspects the preprocessed data and sends unqualified preprocessed data to the manual intervention unit for processing to obtain qualified preprocessed data.

[0016] Adding a manual intervention unit to the data processing module can ensure that all original data can be removed by removing spaces and special values, thereby improving the accuracy of the original data and ensuring the accuracy of subsequent evaluations.

[0017] Furthermore, the quality inspection unit and the manual intervention unit are respectively connected to a protocol conversion unit, which performs protocol conversion and unifies associated fields on the pre-processed data to obtain the standard data, and sends the standard data to the data fusion layer.

[0018] The protocol conversion unit unifies all pre-processed data into the same format, which is convenient for use by the data fusion layer and the MPP data service layer.

[0019] Furthermore, the data fusion layer is provided with a classification module, in which all the standard data generate unique identifiers, and the classification module classifies all the standard data into the enterprise historical project data and enterprise credit comprehensive data according to the identifiers;

[0020] Among them, the enterprise's historical project data is subdivided into the total number of historical projects (a1), the number of successful core business projects (a2), the number of successful non-core business projects (a3), the number of successful innovation projects (a4), the execution cycle of each project (a5), the number of invention patents (a6), the number of utility model patents (a7), and the average contract amount of projects (a8);

[0021] Comprehensive corporate credit data is subdivided into basic corporate information c1, corporate operating information c2, operating risk information c3, and judicial information c4.

[0022] The classification module classifies all standard data according to the identification, which makes it convenient for the enterprise project comprehensive evaluation module and the enterprise credit comprehensive evaluation module to obtain the required calculation data from the enterprise historical project data and the enterprise credit comprehensive data respectively, thereby reducing the calculation amount of the enterprise credit risk prediction system and improving the system's operating speed.

[0023] Furthermore, the enterprise project comprehensive evaluation module is provided with a historical project evaluation unit and a new project feature extraction unit;

[0024] The historical project evaluation unit calculates the enterprise historical project evaluation result a based on the standard data in the enterprise historical project data;

[0025] The calculation formula for the enterprise historical project evaluation result a is:

[0026]

[0027] Among them, a is the evaluation result of the enterprise's historical projects, and a9 is the success rate of historical projects;

[0028] The enterprise's historical project evaluation results a include the total number of historical projects a1, the number of successful main business projects a2, the number of successful non-main business projects a3, the number of successful innovation projects a4, the project execution period a5, the number of invention patents a6, the number of utility model patents a7, the average project contract amount a8, and the historical project success rate a9;

[0029] The new project feature extraction unit receives new project standard data, extracts project features of the new project from the new project standard data, and outputs a new project feature value b. The new project feature value b includes the technical field involved in the new project b1, the proportion of the new project investment amount in the company's net assets b2, and the proportion of the new project investment amount in the investment amount of all currently ongoing projects b3.

[0030] Furthermore, the digital model for comprehensive evaluation of enterprise projects is:

[0031]

[0032] Among them, x is the comprehensive evaluation result of the enterprise project, MATCH is the search function, b1 is the technical field involved in the new project, a is the evaluation result of the enterprise's historical projects, b2 is the proportion of the new project investment amount in the enterprise's net assets, b3 is the proportion of the new project investment amount in the investment amount of all currently carried out projects, a9 is the success rate of historical projects, w1 is the weight of the technical field involved in the new project, w2 is the weight of the proportion of the new project investment amount in the enterprise's net assets, w3 is the weight of the new project investment amount in the investment amount of all currently carried out projects, and w4 is the weight of the historical project success rate.

[0033] Furthermore, the digital model for comprehensive corporate credit evaluation is:

[0034]

[0035] Among them, z is the comprehensive evaluation result of the enterprise credit investigation, c1 is the basic information of the enterprise, c2 is the business information of the enterprise, c3 is the business risk information, c4 is the judicial information, s1 is the weight of the basic information of the enterprise, s2 is the weight of the business information of the enterprise, s3 is the weight of the business risk information, and s4 is the weight of the judicial information.

[0036] Furthermore, the enterprise project credit evaluation digital model is:

[0037]

[0038] Among them, y is the credit evaluation result of the enterprise project, x is the comprehensive evaluation result of the enterprise project, z is the comprehensive evaluation result of the enterprise credit, q1 is the weight of the comprehensive evaluation result of the enterprise project, and q2 is the weight of the comprehensive evaluation result of the enterprise credit.

[0039] A method for predicting corporate credit risk based on big data, the key of which is to include the following steps:

[0040] S1: The data processing module obtains the original data from the data source, processes and verifies it, generates standard data, and sends it to the classification module;

[0041] S11: The preprocessing unit removes spaces and special values from the original data to obtain preprocessed data, and sends the preprocessed data to the quality inspection unit;

[0042] S12: The quality inspection unit performs quality verification on the pre-processed data.

[0043] When the pre-processed data passes the verification, the quality inspection unit sends the qualified pre-processed data to the protocol conversion unit;

[0044] When the pre-processed data fails the verification, the quality inspection unit sends the unqualified pre-processed data to the manual intervention unit;

[0045] S13: The manual intervention unit manually processes the unqualified pre-processed data to obtain qualified pre-processed data, and sends the qualified pre-processed data to the protocol conversion unit;

[0046] S14: The protocol conversion unit performs protocol conversion on the pre-processed data and unifies the associated fields to obtain standard data, and sends the standard data to the classification module;

[0047] S2: The classification module processes all standard data to generate unique identifiers for all standard data. The classification module uses the identifiers to classify all the standard data into enterprise historical project data and enterprise credit comprehensive data.

[0048] S3: The enterprise project comprehensive evaluation module outputs the enterprise project comprehensive evaluation results based on the enterprise's historical project data, and sends the enterprise project comprehensive evaluation results to the enterprise project credit evaluation module;

[0049] S31: The historical project evaluation unit outputs the enterprise historical project evaluation results based on the standard data in the enterprise historical project data;

[0050] S32: The new project feature extraction unit receives the new project standard data, extracts the project features of the new project from the new project standard data, and outputs the new project feature value;

[0051] S33: The enterprise project comprehensive evaluation digital model outputs the enterprise project comprehensive evaluation results based on the enterprise's historical project evaluation results and the new project characteristic values;

[0052] S4: The enterprise credit comprehensive evaluation module outputs the enterprise credit comprehensive evaluation result based on the standard data in the enterprise credit comprehensive data, and sends the enterprise credit comprehensive evaluation result to the enterprise project credit evaluation module;

[0053] S5: The enterprise project credit evaluation module outputs the enterprise project credit evaluation results based on the enterprise project comprehensive evaluation results and the enterprise credit comprehensive evaluation results.

[0054] Beneficial effects: The present invention adds project evaluation to the traditional credit assessment. The project evaluation is mainly based on historical project data, supplemented by new project and current project information, so as to predict the feasibility and future development of the client company's new project. Combined with the credit assessment results of the client company, the client company is evaluated in all aspects, which improves the accuracy and reliability of the evaluation results, thereby reducing the working risks of financial institutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a schematic diagram of an enterprise credit risk prediction system based on big data;

[0056] Figure 2 Schematic diagram of data processing module;

[0057] Figure 3 This is a schematic diagram of the comprehensive evaluation module for enterprise projects;

[0058] Figure 4 The following is a flow chart of a corporate credit risk prediction method based on big data. DETAILED DESCRIPTION

[0059] The specific implementation manner and working principle of the present invention will be further described in detail below with reference to the accompanying drawings.

[0060] like Figure 1 As shown, a corporate credit risk prediction system based on big data is provided with an original data layer, in which a data source and a buffer layer are provided. The data source sends the original data to the buffer layer via FTP, and the buffer layer sends the original data to the data standard layer via data.

[0061] like Figure 1 、 2 As shown, the data standard layer is provided with a data processing module, which is provided with a pre-processing unit, a quality inspection unit, a manual intervention unit and a protocol conversion unit;

[0062] The preprocessing unit performs space removal and special value removal processing on the original data to obtain preprocessed data;

[0063] The quality inspection unit inspects the preprocessed data, sends unqualified preprocessed data to the manual intervention unit, and sends qualified preprocessed data to the protocol conversion unit;

[0064] The manual intervention unit processes the unqualified pre-processed data again to obtain qualified pre-processed data, and sends the qualified pre-processed data to the protocol conversion unit;

[0065] The protocol conversion unit performs protocol conversion and unifies the associated fields of all pre-processed data to obtain standard data with consistent format, and sends all standard data to the data fusion layer.

[0066] like Figure 1 As shown, the data fusion layer is provided with a classification module, which generates a unique identifier for all the standard data through the SHA5 method, and divides all the standard data into the enterprise historical project data and enterprise credit comprehensive data through the identifier, so as to facilitate the acquisition and application of the MPP data service layer;

[0067] Among them, the enterprise's historical project data is subdivided into the total number of historical projects (a1), the number of successful main business projects (a2), the number of successful non-main business projects (a3), the number of successful innovation projects (a4), the execution cycle of each project (a5), the number of invention patents (a6), the number of utility model patents (a7), and the average contract amount of projects (a8);

[0068] Comprehensive corporate credit data is subdivided into basic corporate information c1, corporate operating information c2, operating risk information c3, and judicial information c4.

[0069] like Figure 1 、 Figure 3 As shown, the MPP data service layer is provided with an enterprise project comprehensive evaluation module, an enterprise credit comprehensive evaluation module and an enterprise project credit evaluation module;

[0070] Among them, the enterprise project comprehensive evaluation module is equipped with a historical project evaluation unit, a new project feature extraction unit and an enterprise project comprehensive evaluation digital model;

[0071] The historical project evaluation unit calculates the enterprise historical project evaluation result a based on the standard data in the enterprise historical project data, and sends the enterprise historical project evaluation result a to the enterprise project comprehensive evaluation digital model;

[0072] The calculation formula for the enterprise historical project evaluation result a is:

[0073]

[0074] Among them, a is the evaluation result of the enterprise's historical projects, and a9 is the success rate of historical projects;

[0075] The enterprise's historical project evaluation results a include the total number of historical projects a1, the number of successful main business projects a2, the number of successful non-main business projects a3, the number of successful innovation projects a4, the project execution period a5, the number of invention patents a6, the number of utility model patents a7, the average project contract amount a8, and the historical project success rate a9;

[0076] The new project feature extraction unit receives the new project standard data, extracts the project features of the new project from the new project standard data, and outputs a new project feature value b, which includes the technical field b1 involved in the new project, the proportion of the new project investment amount in the company's net assets b2, and the proportion of the new project investment amount in the investment amount of all currently ongoing projects b3;

[0077] At the same time, the new project feature extraction unit sends the new project feature value b to the enterprise project comprehensive evaluation digital model;

[0078] The enterprise project comprehensive evaluation digital model outputs the enterprise project comprehensive evaluation result x based on the enterprise's historical project evaluation result a and the new project characteristic value b, and sends it to the enterprise project credit evaluation module;

[0079] Among them, the digital model for comprehensive evaluation of enterprise projects is:

[0080]

[0081] Where x is the comprehensive evaluation result of the enterprise project, MATCH is the search function, b1 is the technical field involved in the new project, a is the evaluation result of the enterprise's historical projects, b2 is the proportion of the new project investment amount in the enterprise's net assets, b3 is the proportion of the new project investment amount in the investment amount of all currently ongoing projects, a9 is the success rate of historical projects, w1 is the weight of the technical field involved in the new project, w2 is the weight of the proportion of the new project investment amount in the enterprise's net assets, w3 is the weight of the proportion of the new project investment amount in the investment amount of all currently ongoing projects, and w4 is the weight of the success rate of historical projects;

[0082] The enterprise credit comprehensive evaluation module is provided with an enterprise credit comprehensive evaluation digital model, which outputs the enterprise credit comprehensive evaluation result z based on the standard data in the enterprise credit comprehensive data, and sends the enterprise credit comprehensive evaluation result z to the enterprise project credit evaluation module;

[0083] Among them, the digital model for comprehensive evaluation of corporate credit is:

[0084]

[0085] Among them, z is the comprehensive evaluation result of the enterprise credit investigation, c1 is the basic information of the enterprise, c2 is the business information of the enterprise, c3 is the business risk information, c4 is the judicial information, s1 is the weight of the basic information of the enterprise, s2 is the weight of the business information of the enterprise, s3 is the weight of the business risk information, and s4 is the weight of the judicial information;

[0086] The enterprise project credit evaluation module is provided with an enterprise project credit evaluation digital model, which outputs an enterprise project credit evaluation result y based on the enterprise project comprehensive evaluation result x and the enterprise credit comprehensive evaluation result z;

[0087] Among them, the digital model for enterprise project credit evaluation is:

[0088]

[0089] Among them, y is the credit evaluation result of the enterprise project, x is the comprehensive evaluation result of the enterprise project, z is the comprehensive evaluation result of the enterprise credit, q1 is the weight of the comprehensive evaluation result of the enterprise project, and q2 is the weight of the comprehensive evaluation result of the enterprise credit.

[0090] like Figure 4 As shown, a method for predicting corporate credit risk based on big data includes the following steps:

[0091] S1: The data processing module obtains the original data from the data source, processes and verifies it, generates standard data, and sends it to the classification module;

[0092] S11: The data source sends the original data to the buffer layer via FTP;

[0093] S12: The buffer layer sends the original data to the pre-processing unit through data

[0094] S13: The preprocessing unit removes spaces and special values from the original data to obtain preprocessed data, and sends the preprocessed data to the quality inspection unit;

[0095] For example, if you input "Chongqing Big Data? Company", the pre-processing unit will remove spaces and output "Chongqing Big Data? Company";

[0096] The pre-processing unit takes the input "Chongqing Big Data Company" and outputs "Chongqing Big Data Company" after removing special values.

[0097] S14: The quality inspection unit performs quality verification on the pre-processed data, such as: standardization verification, null value verification, length verification, and specified data verification;

[0098] When the pre-processed data passes the verification, the quality inspection unit sends the qualified pre-processed data to the protocol conversion unit;

[0099] When the pre-processed data fails the verification, the quality inspection unit sends the unqualified pre-processed data to the manual intervention unit;

[0100] S15: The manual intervention unit manually processes the unqualified pre-processed data to obtain qualified pre-processed data, and sends the qualified pre-processed data to the protocol conversion unit;

[0101] S16: The protocol conversion unit performs protocol conversion on the pre-processed data and unifies the associated fields to obtain standard data, and sends the standard data to the classification module;

[0102] S2: The classification module divides the standard data into enterprise historical project data and enterprise credit comprehensive data;

[0103] S21: Determine whether the standard data comes from the designated data source;

[0104] When standard data comes from the designated data source, the standard data is adopted;

[0105] When the standard data does not come from the designated data source, proceed to step S22;

[0106] S22: judging the quality of non-specified data sources;

[0107] When the quality of non-specified data sources is high, the standard data is used;

[0108] When the quality of the non-designated data source is low, proceed to step S23;

[0109] S23: Standard data from low-quality non-specified data sources is adopted after it meets the custom rules;

[0110] S24: All adopted standard data are processed by the SHA5 method, and all adopted standard data are generated with a unique identifier;

[0111] S25: The classification module classifies the business according to the identification;

[0112] S3: The enterprise project comprehensive evaluation module outputs the enterprise project comprehensive evaluation results based on the enterprise's historical project data, and sends the enterprise project comprehensive evaluation results to the enterprise project credit evaluation module;

[0113] S31: The historical project evaluation unit outputs the enterprise historical project evaluation results based on the standard data in the enterprise historical project data;

[0114] S32: The new project feature extraction unit receives the new project standard data, extracts the project features of the new project from the new project standard data, and outputs the new project feature value;

[0115] S33: The enterprise project comprehensive evaluation digital model outputs the enterprise project comprehensive evaluation results based on the enterprise's historical project evaluation results and the new project characteristic values;

[0116] S4: The enterprise credit comprehensive evaluation module outputs the enterprise credit comprehensive evaluation result based on the standard data in the enterprise credit comprehensive data, and sends the enterprise credit comprehensive evaluation result to the enterprise project credit evaluation module;

[0117] S5: The enterprise project credit evaluation module outputs the enterprise project credit evaluation results based on the enterprise project comprehensive evaluation results and the enterprise credit comprehensive evaluation results.

Claims

1. A corporate credit risk prediction system based on big data, including an original data layer, a data standard layer and a data fusion layer, wherein: The data standard layer obtains raw data from the raw data layer, processes the raw data to obtain standard data, and sends the standard data to the data fusion layer. The data fusion layer performs fusion and classification processing on all standard data and classifies the standard data into at least one category. It is characterized by also providing an MPP data service layer, which is provided with an enterprise project comprehensive evaluation module, an enterprise credit comprehensive evaluation module, and an enterprise project credit evaluation module. Among them, the enterprise project comprehensive evaluation module is equipped with an enterprise project comprehensive evaluation digital model, which outputs the enterprise project comprehensive evaluation results. ; The enterprise credit comprehensive evaluation module is equipped with an enterprise credit comprehensive evaluation digital model, which outputs the enterprise credit comprehensive evaluation results. ; The enterprise project credit evaluation module is equipped with an enterprise project credit evaluation digital model, which is based on the comprehensive evaluation results of the enterprise project. and comprehensive evaluation results of corporate credit Output enterprise project credit evaluation results ; The enterprise project comprehensive evaluation module is provided with a historical project evaluation unit and a new project feature extraction unit; Among them, the historical project evaluation unit calculates the enterprise historical project evaluation results based on the standard data in the enterprise historical project data ; The evaluation results of the enterprise's historical projects The calculation formula is: in, The evaluation results of the company's historical projects are is the historical project success rate; The evaluation results of the enterprise's historical projects Including the total number of historical projects , the number of successful main business projects , the number of successful non-core business projects , number of successful innovation projects , project execution cycle , number of invention patents , the number of utility model patents , average contract value of projects and historical project success rates ; The new project feature extraction unit receives the new project standard data, extracts the project features of the new project from the new project standard data, and outputs the new project feature value , the new item characteristic value Includes technical areas involved in new projects 2. The proportion of new project investment amount in the company's net assets and the proportion of new project investment amount in all currently ongoing project investment amounts ; The digital model for comprehensive evaluation of enterprise projects is: in, Comprehensive evaluation results of enterprise projects, To find the function, For the technical fields involved in the new project, The evaluation results of the company's historical projects are is the proportion of new project investment amount in the company's net assets, The proportion of new project investment amount in all currently ongoing project investment amounts, is the historical project success rate, The weight of the technical field involved in the new project, is the weight of the new project investment amount in the company's net assets, The weight of the new project investment amount in the total investment amount of all currently ongoing projects, is the success rate weight of historical projects.

2. The enterprise credit risk prediction system based on big data according to claim 1 is characterized in that: The original data layer includes a data source, which sends the original data to a buffer layer, and the buffer layer transmits the original data to the data standard layer.

3. The enterprise credit risk prediction system based on big data according to claim 1 is characterized in that: The data standard layer is provided with a data processing module, which includes a pre-processing unit and a quality inspection unit; The preprocessing unit obtains the original data and performs space removal and special value removal processing on the original data to obtain preprocessed data; The quality inspection unit is connected to a manual intervention unit, and the quality inspection unit inspects the preprocessed data and sends unqualified preprocessed data to the manual intervention unit for processing to obtain qualified preprocessed data.

4. The enterprise credit risk prediction system based on big data according to claim 3 is characterized in that: The quality inspection unit and the manual intervention unit are respectively connected to a protocol conversion unit, which performs protocol conversion and unifies associated fields on the pre-processed data to obtain the standard data, and sends the standard data to the data fusion layer.

5. The enterprise credit risk prediction system based on big data according to claim 1 is characterized in that: The data fusion layer is provided with a classification module, in which all the standard data generate a unique identifier, and the classification module divides all the standard data into the enterprise historical project data and enterprise credit comprehensive data according to the identifier; Among them, the enterprise historical project data is subdivided into the total number of historical projects , Number of successful main business projects , Number of successful non-core business projects , number of successful innovation projects , Project Execution Cycle , number of invention patents , the number of utility model patents and average contract value of projects ; Comprehensive corporate credit data is subdivided into basic corporate information , business information , business risk information and judicial information .

6. The enterprise credit risk prediction system based on big data according to claim 1, characterized in that: The digital model for comprehensive enterprise credit evaluation is: in, The comprehensive evaluation results of corporate credit investigation are: Basic information of the enterprise, For business information, For business risk information, For judicial information, is the basic information weight of the enterprise, is the weight of enterprise management information, is the operating risk information weight, The weight of judicial information.

7. The enterprise credit risk prediction system based on big data according to claim 1, characterized in that: The digital model for enterprise project credit evaluation is: Among them, y is the credit evaluation result of the enterprise project, The comprehensive evaluation results of enterprise projects are: The comprehensive evaluation results of corporate credit investigation are: is the weight of the comprehensive evaluation results of the enterprise project, The weight of the comprehensive evaluation results of corporate credit investigation.

8. A method for predicting corporate credit risk based on big data, applied to a system for predicting corporate credit risk based on big data as claimed in any one of claims 1 to 7, characterized in that: The following steps are involved: S1: The data processing module obtains the original data from the data source, processes and verifies it, generates standard data, and sends it to the classification module; S11: The preprocessing unit removes spaces and special values from the original data to obtain preprocessed data, and sends the preprocessed data to the quality inspection unit; S12: The quality inspection unit performs quality verification on the pre-processed data. When the pre-processed data passes the verification, the quality inspection unit sends the qualified pre-processed data to the protocol conversion unit; When the pre-processed data fails the verification, the quality inspection unit sends the unqualified pre-processed data to the manual intervention unit; S13: The manual intervention unit manually processes the unqualified pre-processed data to obtain qualified pre-processed data, and sends the qualified pre-processed data to the protocol conversion unit; S14: The protocol conversion unit performs protocol conversion on the pre-processed data and unifies the associated fields to obtain standard data, and sends the standard data to the classification module; S2: The classification module processes all standard data to generate unique identifiers for all standard data. The classification module uses the identifiers to classify all the standard data into enterprise historical project data and enterprise credit comprehensive data. S3: The enterprise project comprehensive evaluation module outputs the enterprise project comprehensive evaluation results based on the enterprise's historical project data, and sends the enterprise project comprehensive evaluation results to the enterprise project credit evaluation module; S31: The historical project evaluation unit outputs the enterprise historical project evaluation results based on the standard data in the enterprise historical project data; S32: The new project feature extraction unit receives the new project standard data, extracts the project features of the new project from the new project standard data, and outputs the new project feature value; S33: The enterprise project comprehensive evaluation digital model outputs the enterprise project comprehensive evaluation results based on the enterprise's historical project evaluation results and the new project characteristic values; S4: The enterprise credit comprehensive evaluation module outputs the enterprise credit comprehensive evaluation result based on the standard data in the enterprise credit comprehensive data, and sends the enterprise credit comprehensive evaluation result to the enterprise project credit evaluation module; S5: The enterprise project credit evaluation module outputs the enterprise project credit evaluation results based on the enterprise project comprehensive evaluation results and the enterprise credit comprehensive evaluation results.

Images