Public loan system and computer storage medium

Abstract

The invention discloses a peer-to-peer loan system and a computer storage medium, and relates to the technical field of computers, and the peer-to-peer loan system comprises a customer group management module, an intelligent auditing module, a credit evaluation module and a whole-process management and control module. The customer group management module divides the to-be-classified enterprises into different enterprise groups according to the multi-dimensional features, the multi-dimensional features are adopted for division, the fine degree of enterprise division is increased, and the precision of public loan service is further improved; the intelligent auditing module provides a basic screening rule and a personalized screening rule of each enterprise group, and further adopts the personalized screening rule to carry out enterprise screening on the basis of basic screening, so that the quality of enterprises with loan qualification is improved, and the service quality of public loan service can also be ensured; the credit evaluation module is specifically used for calculating a pre-credit line; and the full-process management and control module is used for monitoring the handling process of the credit extension service, performing full-process tracing and multi-stage rechecking, and further improving the service quality and accuracy of the public loan service.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a corporate loan system and a computer storage medium. Background Art

[0002] With the deep integration of HP Finance and enterprise services, commercial banks are accelerating the transformation of digital credit services. Currently, enterprise loan services generally utilize a standardized review process. However, this process still has problems: the traditional enterprise loan service model relies on a unified evaluation system that cannot accurately adapt to the differentiated needs of enterprises of different industries and sizes, resulting in low efficiency in identifying high-quality customers. Furthermore, the accuracy and automation level of enterprise loan services provided by existing technologies are relatively low. Summary of the Invention

[0003] In view of the above problems, this application provides a corporate loan system and computer storage medium to achieve the purpose of providing a corporate loan service with high accuracy and automation. The specific solution is as follows:

[0004] The first aspect of this application provides a corporate loan system, including a customer management module, an intelligent review module, a credit assessment module, and a full-process control module:

[0005] The customer group management module is used to divide the enterprises to be classified into different enterprise groups according to multi-dimensional characteristics;

[0006] An intelligent review module is used to screen out, from different enterprise groups, a first enterprise group that meets the basic screening rules and the personalized screening rules pre-configured for the different enterprise groups; wherein one enterprise group corresponds to one personalized screening rule;

[0007] a credit assessment module, configured to calculate a pre-credit limit for the first enterprise group based on a credit limit assessment model;

[0008] The full-process management and control module is used to handle credit business for any enterprise in the first enterprise group, generate a business process tracing report and a multi-level review task list.

[0009] In one possible implementation, the multi-dimensional features include industry classification, scale classification, and credit rating. The customer group management module is specifically used to:

[0010] The customer group management module divides the enterprises to be classified into different enterprise groups according to industry classification, scale classification and credit rating.

[0011] In one possible implementation, the customer group management module includes an enterprise classification model and a dynamic tag management model. The customer group management module is specifically used to:

[0012] The enterprises to be classified are input into the enterprise classification model and dynamic tag management model in turn to obtain various enterprise groups with enterprise group labels; among them, one enterprise group corresponds to a unique enterprise group label; the enterprise group label includes the industry to which it belongs, scale and credit rating; the enterprise classification model and dynamic tag management model are both constructed based on the K-means clustering algorithm.

[0013] In one possible implementation, the intelligent audit module is specifically used to:

[0014] The intelligent audit module runs on a rule engine, which selects a second group of companies that meet the basic screening rules from different groups of companies;

[0015] The rule engine selects the first enterprise group that meets the personalized screening rules from the second enterprise group according to the personalized screening rules corresponding to each enterprise group in the second enterprise group.

[0016] In one possible implementation, the intelligent audit module also includes a dynamic threshold adjustment algorithm;

[0017] The intelligent audit module adopts a dynamic threshold adjustment algorithm to dynamically adjust the rule settings in the basic screening rules and personalized screening rules based on external environmental conditions.

[0018] In one possible implementation, the credit limit assessment model is an extreme gradient boosting credit scoring model that incorporates Monte Carlo simulation credit limit calculation. The credit limit assessment module is specifically used to:

[0019] Extracting feature engineering from the enterprise data of the first enterprise group based on the extreme gradient boosting credit scoring model, performing feature importance assessment for the feature engineering, obtaining feature importance assessment results, and further performing credit scoring for each enterprise group in the first enterprise group based on the feature importance assessment results to obtain a credit limit for each enterprise group;

[0020] Monte Carlo simulation credit limit calculation is used to dynamically simulate the revenue scenarios of various enterprise groups to obtain the risk data of each enterprise group. Based on the risk data, the credit limit output in the extreme gradient improvement credit scoring model is optimized to obtain the optimized pre-credit limit.

[0021] In one possible implementation, the full-process control module handles a credit business for any enterprise in the first enterprise group and is specifically configured as follows:

[0022] The full-process management and control module uses blockchain evidence storage technology and a workflow engine to monitor the credit business processing of any enterprise in the first enterprise group.

[0023] In one possible implementation, the full-process control module handles a credit business for any enterprise in the first enterprise group and is specifically configured as follows:

[0024] Generate a unique application number for any enterprise;

[0025] Record the operation logs of any enterprise during the credit business process and generate process tracing reports based on the operation logs;

[0026] During the credit business processing, multi-level review tasks are performed and a multi-level review task table is generated.

[0027] In a possible implementation, the corporate loan system further includes an output module;

[0028] The output module is used to output the pre-credit result file, risk warning report and task list of the first enterprise group.

[0029] The second aspect of the present application provides a computer storage medium, which carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement the corporate loan system of the above-mentioned first aspect or any implementation method of the first aspect.

[0030] By means of the above technical solution, the corporate loan system and computer storage medium provided by this application include a customer group management module, an intelligent review module, a credit assessment module and a full-process control module. The customer group management module divides the enterprises to be classified into different enterprise groups according to multi-dimensional characteristics. The multi-dimensional characteristics are used for division, which increases the precision of enterprise division and further improves the accuracy of corporate loan services; the intelligent review module provides basic screening rules and personalized screening rules for each enterprise group. On the basis of basic screening, personalized screening rules are further used to screen enterprises, thereby improving the quality of enterprises with loan qualifications and ensuring the service quality of corporate loan services; the credit assessment module is specifically used to calculate the pre-credit limit; the full-process control module is used to monitor the processing process of credit business, conduct full-process tracing and multi-level review, and further improve the service quality and accuracy of corporate loan services. The corporate loan system provided by this application can provide fully automatic, high-quality and high-precision corporate loan services based on the above modules. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The above and other features, advantages, and aspects of the various embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that the originals and elements are not necessarily drawn to scale.

[0032] Figure 1 A schematic diagram of the structure of the corporate loan system provided for this application;

[0033] Figure 2An example diagram of the structure of the corporate loan system provided for this application. DETAILED DESCRIPTION

[0034] The following describes the embodiments of the present application in conjunction with the accompanying drawings. The terms used in the implementation methods of the present application are only used to explain the specific embodiments of the present application and are not intended to limit the present application.

[0035] The embodiments of the present application are described below in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0036] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, and this is merely a way of distinguishing the objects of the same attributes when describing them in the embodiments of the present application. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, so that the process, method, system, product or equipment comprising a series of units need not be limited to those units, but may include other units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0037] The standardized review process used in existing corporate loan services still has the following problems:

[0038] First: The review rules of traditional corporate loan services are rigid and lack dynamic adjustment capabilities. They are unable to cope with real-time fluctuations in corporate operating data, such as changes in corporate revenue and policy adjustments, which affects the accuracy of customer identification.

[0039] Second: Risk assessment in traditional corporate loan services mostly relies on static financial indicators and cannot effectively integrate multi-dimensional data such as supply chain, taxation, and public opinion, which affects the accuracy of credit limit calculation.

[0040] Third: The business processes of traditional corporate loan services have information silos. When providing services, account managers need to manually handle repetitive tasks such as customer list screening and historical business verification across systems, which can easily lead to operational risks and compliance risks.

[0041] In summary, the accuracy and degree of automation of the corporate loan services provided by existing technologies are relatively low.

[0042] To solve the above problems, existing technologies use enterprise whitelist management and automated approval tools. However, there are still many problems in actual applications, which makes it difficult for enterprises to achieve efficient and accurate risk management and business collaboration.

[0043] Problem 1: Extensive customer segmentation: Single-dimension segmentation leads to resource mismatch.

[0044] Existing tools mostly rely on basic industry classifications (such as manufacturing and finance) or single financial indicators (such as asset size and income level) to divide customer groups, and lack multi-dimensional dynamic analysis capabilities.

[0045] For example, a commercial bank classified its customers into A / B / C categories based solely on the company's annual revenue, resulting in a large number of "hidden high-quality customers" (such as technology companies with high R&D investment but short-term losses) being classified as low-value customers and missing out on credit opportunities.

[0046] A certain e-commerce platform only divides users into levels based on historical consumption amounts, without incorporating behavioral data such as user activity and repurchase rate. This results in marketing resources being wasted on the "low-frequency, high-consumption" customer group, while high-potential "high-frequency, low-consumption" customers are not effectively activated.

[0047] This demonstrates that extensive stratification leads to discrepancies in the identification of high-value customers, and a shift in marketing and risk control resources toward less-efficient customer groups. For example, a manufacturing company allocated 80% of its credit line to traditional industry clients, while neglecting asset-light but high-growth SMEs in the new energy industry chain, thus missing out on the benefits of industry transformation.

[0048] Moreover, extensive stratification easily fails to identify changes in the customer life cycle. For example, a financial institution failed to incorporate temporary revenue fluctuations of medical supplies manufacturers during the epidemic into the stratification model, resulting in the loss of high-quality customers due to temporary downgrades.

[0049] Problem 2: Rigid rule configuration: Policy response lags behind the high cost of manual intervention.

[0050] Existing system rule bases are mostly static, based on historical experience, making them difficult to adapt to dynamic policy changes and specific business needs. This is mainly manifested in rigid exemption clauses and frequent manual intervention.

[0051] Regarding the rigid exemption clauses, for example, a local government's green industry subsidy policy requires the relaxation of credit conditions for environmental protection technology companies, but a bank's whitelist system still uses traditional financial indicators (such as asset-liability ratio ≤ 70%), resulting in a number of innovative companies that meet policy guidelines but have excessive short-term debt ratios being mistakenly rejected.

[0052] For specialized and innovative enterprises with frequent manual intervention, the value of their technical patents is difficult to quantify, but the existing system lacks an intellectual property evaluation module, and manual submission of materials for review is required one by one, which extends the approval cycle by 3-5 days and affects the financing efficiency of the enterprise.

[0053] In addition, existing corporate loan services have low rule coupling, and data silos hinder dynamic adjustments.

[0054] There are conflicts between exemption requirements for different policy scenarios (e.g., carbon neutrality and rural revitalization). For example, a cross-border e-commerce platform needs to meet both "support for small and medium-sized enterprises" and "anti-money laundering" regulations, but the system cannot automatically identify the overlap in compliance, requiring manual compilation of an additional list.

[0055] Data silos hinder dynamic adjustments: policy documents, industry whitelists, and real-time business data of enterprises are scattered across multiple systems. The rule engine is unable to obtain external data in real time, resulting in a delay of several months in the update of exemption clauses.

[0056] Question 3: Lack of a full-cycle risk control mechanism: Risk control gaps and lags are prominent.

[0057] Existing corporate loan services focus on pre-loan approval and lack dynamic monitoring of credit behavior during the loan and post-loan risks.

[0058] A lack of monitoring during loan issuance manifested itself in the following ways: a supply chain finance platform only reviewed and approved transaction contracts at the time of credit approval, failing to monitor the flow of funds. A company illegally used a loan for real estate speculation, and the loan service provider failed to detect the anomaly through transaction flow analysis, ultimately resulting in a bad debt.

[0059] Post-loan early warning lags manifested themselves in the following ways: One consumer finance company relied on quarterly manual post-loan checks and failed to access real-time data such as tax and social security data. A restaurant company suffered sudden losses due to the pandemic, but the company's loan service system failed to trigger risk signals in a timely manner, leading to a surge in delinquency rates.

[0060] It can be seen that there are many problems with corporate loan services in existing technologies, which cause banks to face severe challenges in various aspects such as optimizing customer experience, improving service efficiency and risk management.

[0061] In order to solve the above problems, this application provides a corporate loan system.

[0062] Optional, see Figure 1 , a structural diagram of the corporate loan system provided in this application.

[0063] like Figure 1 As shown, the corporate loan system includes a customer management module, an intelligent review module, a credit assessment module and a full-process supervision module.

[0064] Among them, the customer group management module is specifically used to automatically stratify (classify) the enterprises to be classified and divide the enterprises to be classified into different enterprise groups.

[0065] The intelligent audit module is specifically used to audit whether different classified enterprise groups have credit qualifications, and to select the first enterprise group with credit qualifications from the different enterprise groups.

[0066] Next, the credit module is used to calculate the pre-credit limit for the first enterprise group.

[0067] The full-process control module is used to monitor the credit business processing process of any enterprise in the first enterprise group, and generate a business processing process tracing report and a multi-level review task table.

[0068] Optionally, the customer group management module divides the enterprises to be classified into different enterprise groups according to multi-dimensional characteristics.

[0069] Existing technologies generally use coarse-grained classification features, such as using the major categories of national standard industry classification (agriculture, forestry, animal husbandry and fishery, manufacturing, etc.) to classify enterprises. Extensive classification can easily lead to biased identification of high-value customers, and marketing and risk control resources will be tilted towards inefficient customer groups. In addition, it will also affect the accuracy of corporate loan services.

[0070] In this application, multi-dimensional features are used as the criteria for dividing corporate customer groups. Multi-dimensional features are refined classification features set from different dimensions.

[0071] For example, the multi-dimensional features provided in this application are based on counties and are divided by enterprise size, sales revenue, and smaller-granularity national standard industry classification features. For example: Qinghe cashmere, Ningjin cable, Tangxian sheep farming.

[0072] Optional multi-dimensional features include industry classification, scale classification, and credit rating. The customer group management module divides the classified enterprises into different enterprise groups based on industry classification, national model classification, and credit rating. Industry classification is primarily based on industry standard codes matched to the enterprise's main business keywords; scale classification is primarily based on registered resources / annual revenue, using a dynamic quantile algorithm to determine enterprise size; and credit rating is the latest enterprise credit rating recorded in each enterprise's data.

[0073] Specifically, the customer group management module includes a business classification model and a dynamic tag management model, both of which are built based on the K-means clustering algorithm. The businesses to be classified are sequentially input into the business classification model and the dynamic tag management model, resulting in individual business groups with enterprise group labels. Each business group corresponds to a unique enterprise group label, which generally consists of: industry + company size + credit rating. For example: manufacturing - small and medium-sized - good credit, technology-based enterprises - high potential - excellent credit.

[0074] Next, we will introduce the K-means clustering algorithm. The core principle of this algorithm for classifying companies is to quantify industry characteristics and divide industries into several groups based on similarity. The classification process can be divided into the following steps:

[0075] Step 1. Data preparation: quantification of industry characteristics.

[0076] First, perform feature selection: select quantitative indicators that reflect the essence of the industry. Examples include financial indicators, market indicators, operational indicators, and external environment indicators. Financial indicators include revenue growth rate, profit margin, and debt-to-asset ratio; market indicators include market size, market share, and user growth rate; operational indicators include R&D investment ratio, supply chain complexity, and technology dependency; and external environment indicators include policy sensitivity and cyclical fluctuations.

[0077] Then, perform data standardization: Given the dimensional differences of different indicators (such as revenue in "billion yuan" and profit margin in "percentage"), it is necessary to eliminate the dimensional impact through standardization (such as Z-score) to ensure that the weights of each dimension are balanced.

[0078] Step 2: Determine the number of clusters (K value).

[0079] Methods include empirical methods, elbow rules, and contour systems.

[0080] Among them, the empirical method is pre-set based on industry knowledge, such as dividing the industrial chain into upstream, midstream and downstream;

[0081] The elbow rule is to calculate the total sum of squares (SST) under different K values ​​and select the K value corresponding to the inflection point of SST decline;

[0082] The contour system is used to evaluate the closeness of samples within their categories and the separation between categories, and the K value that maximizes the coefficient is selected.

[0083] Step 3: Initialize the cluster center.

[0084] Randomly select K industries as the initial cluster centers, or optimize the initial center selection through K-means+ to avoid falling into local optimality.

[0085] Step 4: Iterative allocation and update.

[0086] First, the allocation steps:

[0087] Each industry sample is assigned to the cluster center with the nearest distance, and the Euclidean distance is usually used to measure the similarity.

[0088] Then, the update step: recalculate the center point of each cluster, that is, the mean of all industry feature vectors in the cluster.

[0089] Finally, the process terminates when the termination condition is met. It stops when the cluster center no longer changes significantly or when the maximum number of iterations is reached.

[0090] Step 5. Industry classification results.

[0091] Finally, K clusters are obtained. The industries within each cluster are highly similar in feature space, while the differences between different clusters are significant.

[0092] For example, Cluster 1: customer-oriented enterprises with high growth and high R&D investment, such as artificial intelligence and biomedicine; Cluster 2: traditional industries with low growth and high stability, such as utilities and basic chemicals; Cluster 3: industries with strong cyclicality and policy sensitivity, such as real estate and non-ferrous metals.

[0093] Step 6: Verify and interpret the results.

[0094] First, internal validation was performed to evaluate cluster tightness and separation using the silhouette coefficient and Davies-Bouldin index;

[0095] Then perform external verification, if there is labeled data, calculate the accuracy, F1 score, etc.

[0096] Finally, the business explanation is given, and the common characteristics of each category are analyzed in combination with industry knowledge, and the scores are given practical meanings, such as "technology-driven" and "resource-dependent".

[0097] Next, we will introduce the input and output of the customer group management module in practical applications.

[0098] The customer management module is specifically used to implement multi-dimensional classification and list management of enterprise groups. The input is the Excel / CSV list of enterprises to be classified imported from the outside, and the industry database called through the API interface; the output is different enterprise groups with enterprise group labels after stratification.

[0099] It should also be noted that the corporate loan system undergoes a data collection phase before importing the Excel / CSV list of companies to be classified and accessing industry databases through APIs. The collected data primarily includes external and internal data sources. External data sources (company lists, tax data) can be imported through standardized interfaces, while internal data (customer declaration information, historical credit records) can be synchronized from the credit center system.

[0100] Specifically, it mainly imports an Excel / CSV list of companies to be classified from the outside and parses the content of the list into structured data; and calls the industry database through the API interface. The API interface can be a tax / industrial and commercial data interface, and the data in json format is fed back through the interface; in addition, the financial statements submitted by each company and historical credit records are queried from the internal database of the system.

[0101] The above-mentioned customer group management module divides the enterprises to be classified into different enterprise groups based on refined multi-dimensional features, supports batch import of external lists, and solves the problem of extensive customer stratification in traditional corporate loan services.

[0102] Next, we will introduce the intelligent review module, which mainly performs two-level screening. After obtaining the classification results from the customer management module, it matches and screens the rules to select the first enterprise group from different enterprise groups that meets both the basic screening rules and the personalized screening rules of each enterprise group.

[0103] This module mainly runs based on a rule engine, which first selects a second enterprise group that meets the basic screening rules from different enterprise groups, and then selects a first enterprise group that meets its personalized screening rules from the second enterprise group according to the personalized screening rules corresponding to each enterprise group in the second enterprise group.

[0104] Among them, the basic screening rules can be understood as a preset universal threshold, which is a basic standard that all credit-granting enterprises must achieve, and is also a mandatory verification rule; each enterprise group in the second enterprise group will correspond to a personalized screening rule. Personalized screening rules can be understood as different verification rules set for different customer groups. For example, Ningjin Cable sets an annual minimum tax value and screens out enterprises that are below the minimum tax value.

[0105] In addition, the intelligent audit module also adopts a dynamic threshold adjustment algorithm to dynamically adjust the rule settings in the basic screening rules and personalized screening rules based on external environmental conditions.

[0106] For example, the rule engine first loads basic screening rules, such as requirements for establishment ≥ 2 years and tax rating B or higher. It then loads customized screening rules for different enterprise groups, such as exemptions for high-tech enterprises from mortgage requirements. Finally, it dynamically adjusts the basic and / or customized screening rules based on external environmental factors, including but not limited to the industry prosperity index. For example, the revenue threshold in the rules can be modified based on the industry prosperity index. Since the industry prosperity index may be affected by economic or policy factors, the revenue threshold needs to be dynamically adjusted based on the changing industry prosperity index.

[0107] Next, we introduce the input and output of the intelligent audit module in practical applications.

[0108] The intelligent audit module is specifically used to execute preset rules and dynamically adjust parameters. Its input is the customer stratification results in the customer management module, the financial statements submitted by the enterprise itself, and the tax / industrial and commercial data captured by the system; the output is the pass / reject audit results and the whitelist exemption mark.

[0109] The above-mentioned intelligent audit module solves the problem of rigid rules in existing corporate loan services through preset basic screening rules and personalized screening rules. It can realize whitelist exemptions and dynamic adjustment of revenue / tax parameters, realize flexible configuration of audit rules, and improve the efficiency of special policy impact by more than 50%.

[0110] Next, the credit assessment module is introduced. This module mainly integrates the audit results of the intelligent audit module with external data, and calculates the pre-credit limit of the first enterprise group through the credit limit assessment model.

[0111] Optionally, the credit limit assessment model provided in this application is an extreme gradient boosting credit scoring model that incorporates Monte Carlo simulation credit limit calculation. Based on the extreme gradient boosting credit scoring model, feature engineering is extracted from the enterprise data of the first enterprise group, feature importance assessment is performed for the feature engineering, and feature importance assessment results are obtained. Further, based on the feature importance assessment results, credit scores are assigned to each enterprise group within the first enterprise group to obtain credit limits for each enterprise group.

[0112] Specifically, the combination of the Extreme Gradient Boosting credit scoring model (XGBoost credit scoring model) and Monte Carlo simulation can optimize credit decisions through quantified risks and dynamic simulation.

[0113] First, data preparation and model training are performed.

[0114] Integrate internal customer data (such as transaction records, repayment history) and external data (such as credit reports, industrial and commercial information);

[0115] Then, key information is extracted from it, and high-discrimination variables are screened through XGBoost’s feature importance evaluation; XGBoost is used to perform credit scoring on the enterprise and output the default probability or risk level (such as low risk, high risk, medium risk).

[0116] Definition of key variables using Monte Carlo simulation.

[0117] Specifically, the risk score output by XGBoost is used as the core variable, and other uncertain factors (such as economic growth rate, industry fluctuations, interest rate changes, etc.) are incorporated; probability distributions are set for variables; scenarios are simulated to generate credit limit estimates, and tens of thousands of combinations of economic environment and corporate behavior are randomly sampled through the Monte Carlo method; then, a credit limit sensitivity analysis is performed: for each simulated scenario, the maximum tolerable credit limit is calculated based on the corporate risk score and income level; finally, the simulation results are summarized to calculate the expected loss, default probability and risk value under different credit limits; the acceptable maximum loss threshold is determined based on risk preference, and the optimal credit limit range is output.

[0118] In summary, this application employs an extreme gradient boosting credit scoring model that incorporates Monte Carlo simulation for credit limit calculation. This model captures static feature relationships through XGBoost and employs Monte Carlo simulation for dynamic environmental changes, enhancing the robustness of the model's credit limit decisions. Furthermore, the model outputs recommended credit limit values ​​with a confidence level of ≥90%. Furthermore, the model integrates enterprise declaration data with external data sources such as tax, industrial and commercial, and judicial sources, enabling intelligent calculation of pre-credit limits and limiting the assessment error rate to less than 5%.

[0119] Next, we introduce the input and output of the credit assessment module in practical applications.

[0120] This module is specifically used to calculate the credit limit of the first enterprise group and generate a risk assessment report.

[0121] The input is the approved enterprise data and the enterprise's external credit data (such as central bank credit / third-party data); the output is the enterprise's pre-credit limit and risk level.

[0122] Finally, the full-process management and control module is introduced, which uses blockchain evidence storage technology and a workflow engine to monitor the credit business processing of any enterprise in the first enterprise group.

[0123] Specifically, a unique application number is generated for any enterprise, which can be in UUID format; the operation log of any enterprise in the process of handling credit business is recorded, and the operation log mainly includes the account manager's operation time and modification content, etc., and a process tracing report is generated based on the operation log; multi-level review tasks are performed in the process of handling credit business, and a multi-level review task table is generated, which includes preliminary review content, risk control review content and final decision content.

[0124] Next, we will introduce the input and output of the full-process control module in actual applications.

[0125] This module is used to manage the full-cycle operation records and review processes of the business. The input is credit application data and historical operation records, and the output is process tracing reports and review task allocation tables.

[0126] In summary, the full-process management and control module can trace historical business records, perform multi-level reviews, and generate visual progress tracking, forming an unalterable audit trail, reducing manual intervention by 70%, and reducing operational risks and compliance risks; and, through a phased verification mechanism (initial review → detailed review → final review) and scheduled task scheduling functions, it shortens the business processing cycle to 1 / 3 of the traditional model, while improving the marketing accuracy and service experience of account managers.

[0127] In addition to the aforementioned modules, the corporate loan system provided in this application also includes an output module that outputs a pre-credit approval result file for the first group of enterprises, a risk warning report (generated when a threshold alarm is triggered), and a task list (a branch task list, automatically divided and sent according to the enterprise within its jurisdiction). The output results are then sent to the branch's CRM (Customer Relationship Management) system via the bank's internal message queue.

[0128] Specifically, the output may be a PDF report containing the amount, term, and interest rate.

[0129] In summary, the corporate loan system provided by this application includes a customer group management module, an intelligent review module, a credit assessment module and a full-process control module, wherein the customer group management module divides the enterprises to be classified into different enterprise groups according to multi-dimensional characteristics, and adopts multi-dimensional characteristics for division, which increases the precision of enterprise division and further improves the accuracy of corporate loan services; the intelligent review module provides basic screening rules and personalized screening rules for each enterprise group, and further adopts personalized screening rules to screen enterprises on the basis of basic screening, thereby improving the quality of enterprises with loan qualifications and ensuring the service quality of corporate loan services; the credit assessment module is specifically used to calculate the pre-credit limit; the full-process control module is used to monitor the processing process of credit business, conduct full-process tracing and multi-level review, and further improve the service quality and accuracy of corporate loan services. The corporate loan system provided by this application can provide fully automatic, high-quality and high-precision corporate loan services based on the above modules.

[0130] For example, see Figure 2 , an example diagram of the structural composition of the corporate loan system provided in this application.

[0131] like Figure 2 As shown, the system mainly includes four parts: system authority role management module, customer group management module, intelligent review module and system business support module.

[0132] Among them, the system authority role management module mainly carries out personnel management, authority management, menu management, organization management and approver management.

[0133] The customer group management module obtains industry databases and manual data from the outside and classifies corporate groups based on various rule configurations.

[0134] The intelligent audit module obtains third-party data related to the enterprise from the outside and then performs two tasks: business acceptance and business investigation. Business acceptance mainly involves data acquisition and then qualification verification; business investigation mainly involves credit calculation based on enterprise + business owner information, production and operation information, and credit mortgage information, followed by business review and finally business confirmation.

[0135] The system business support module mainly explains that the corporate loan system provided by this application supports pre-credit limit re-testing, process management, material management, loan renewal management, whitelist management and other businesses.

[0136] To sum up, the corporate loan system provided in this application can achieve the comprehensive goals of precise corporate loan services, pre-positioned risk management and intelligent business processes, and provide financial institutions with a digital transformation system that can be implemented on a large scale.

[0137] A computer storage medium is also provided in an embodiment of the present application. The storage medium carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement any one of the corporate loan systems provided in the embodiment of the present application.

[0138] It should also be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided in this application, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines.

[0139] Through the description of the above embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software plus necessary general hardware, and of course can also be implemented by special hardware including application-specific integrated circuits, special CPUs, special memories, special components, etc. In general, all functions performed by computer programs can be easily implemented with corresponding hardware, and the specific hardware structures used to implement the same function can also be diverse, such as analog circuits, digital circuits or special circuits, etc. However, for the present application, software program implementation is a better implementation method in most cases. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a readable storage medium, such as a computer's floppy disk, USB flash drive, mobile hard disk, ROM, RAM, magnetic disk or optical disk, etc., and includes a number of instructions to enable a computer device (which can be a personal computer, training equipment, or network equipment, etc.) to execute the methods described in each embodiment of the present application.

[0140] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware, or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0141] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, a computer, a training device or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website, a computer, a training device or a data center. The computer-readable storage medium can be any available medium that a computer can store or a data storage device such as a training device, a data center, etc. that includes one or more available media integrations. The available medium can be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

Claims

1. A corporate loan system, characterized in that: It includes customer group management module, intelligent review module, credit assessment module and full-process control module: The customer group management module is used to divide the enterprises to be classified into different enterprise groups according to multi-dimensional characteristics; The intelligent audit module is used to screen out from the different enterprise groups a first enterprise group that meets the basic screening rules and the personalized screening rules pre-configured for the different enterprise groups; wherein one enterprise group corresponds to one personalized screening rule; The credit assessment module is configured to calculate the pre-credit limit of the first enterprise group based on the credit limit assessment model; The full-process management and control module is used to handle credit business for any enterprise in the first enterprise group and generate a business process tracing report and a multi-level review task table.

2. The corporate loan system according to claim 1, characterized in that: The multi-dimensional features include industry classification, scale classification and credit rating. The customer group management module is specifically used to: The customer group management module divides the enterprises to be classified into the different enterprise groups according to the industry classification, scale classification and credit rating.

3. The corporate loan system according to claim 2, characterized in that: The customer group management module includes an enterprise classification model and a dynamic tag management model. The customer group management module is specifically used to: The enterprises to be classified are sequentially input into the enterprise classification model and the dynamic tag management model to obtain enterprise groups with enterprise group labels; wherein, one enterprise group corresponds to a unique enterprise group label; the enterprise group label includes the industry to which it belongs, the scale and the credit rating; the enterprise classification model and the dynamic tag management model are both constructed based on the K-means clustering algorithm.

4. The corporate loan system according to claim 1, characterized in that: The intelligent audit module is specifically used to: The intelligent audit module runs based on a rule engine, and the rule engine selects a second enterprise group that meets the basic screening rules from the different enterprise groups; The rule engine selects the first enterprise groups that meet the personalized screening rules from the second enterprise groups according to the personalized screening rules corresponding to each enterprise group in the second enterprise group.

5. The corporate loan system according to claim 1, characterized in that: The intelligent audit module also includes a dynamic threshold adjustment algorithm; The intelligent audit module adopts the dynamic threshold adjustment algorithm to dynamically adjust the rule settings in the basic screening rules and the personalized screening rules based on external environmental conditions.

6. The corporate loan system according to claim 1, characterized in that: The credit limit assessment model is an extreme gradient boosting credit scoring model that integrates Monte Carlo simulation credit limit calculation; the credit assessment module is specifically used to: Extracting feature engineering from the enterprise data of the first enterprise group based on the extreme gradient boosting credit scoring model, performing feature importance assessment on the feature engineering to obtain feature importance assessment results, and further performing credit scoring for each enterprise group in the first enterprise group based on the feature importance assessment results to obtain a credit limit for each enterprise group; The Monte Carlo simulation credit limit calculation is used to dynamically simulate the revenue scenarios of the various enterprise groups to obtain the risk data of the various enterprise groups. The credit limit output in the extreme gradient improvement credit scoring model is optimized based on the risk data to obtain the optimized pre-credit limit.

7. The corporate loan system according to claim 1, characterized in that: The full-process control module handles credit business for any enterprise in the first enterprise group and is specifically configured as follows: The full-process management and control module uses blockchain evidence storage technology and a workflow engine to monitor the credit business processing of any enterprise in the first enterprise group.

8. The corporate loan system according to claim 1, characterized in that: The full-process control module handles credit business for any enterprise in the first enterprise group and is specifically configured as follows: Generate a unique application number for any of the aforementioned enterprises; Recording the operation logs of any enterprise in the process of handling credit business, and generating the process tracing report based on the operation logs; During the credit business processing, multi-level review tasks are performed to generate the multi-level review task table.

9. The corporate loan system according to claim 1, characterized in that: The public loan system further includes an output module; The output module is used to output the pre-credit result file, risk warning report and task list of the first enterprise group.

10. A computer storage medium, characterized in that The storage medium carries one or more computer programs, which, when executed by an electronic device, enable the electronic device to implement the corporate loan system as described in any one of claims 1 to 9.

Images