Credit risk early warning method and device, equipment and storage medium

Abstract

The invention discloses a credit risk early warning method, apparatus and device, and a storage medium, and is applied to the field of financial science and technology. Identifying the credit and loan archive data through an information identification model to obtain structured key element information; determining a risk early warning customer according to the key element information and the customer repayment business process information; and according to the risk early warning customer and the customer association information network, predicting a potential association risk customer, and according to the risk early warning customer and the association risk customer, carrying out credit risk early warning. Structured key element information is automatically recognized from credit archive data through the information recognition model, so that the labor cost is saved, the recognition precision is improved, potential associated risk customers can be predicted based on the obtained risk early warning customers and the customer associated information network, and therefore, prospective risk early warning is realized, and the risk early warning efficiency is improved. Business personnel can take effective measures in time to prevent or resolve risks, and the credit asset safety of financial institutions can be guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of financial technology, and in particular to a credit risk early warning method, device, equipment and storage medium. Background Art

[0002] In the process of banking business handling, a large amount of archival data will be generated and accumulated. It is mainly based on the risk and benefit assessment of the internal business experts, and the various prerequisite requirements and supervision norms agreed upon during the business handling and business continuation period, which are directly related to the results of the business approval process and even the management and monitoring during the continuation period.

[0003] At present, post-loan risk warnings for traditional banking credit projects still mainly rely on risk monitoring personnel to conduct targeted monitoring of key products or key customer groups, manually extract key elements of archival materials generated during business processing, organize them into structured information, and then complete offline collision processing of the above structured data with business process data, or manually import them into the system for online matching, and issue risk warnings based on the processing or matching results.

[0004] However, the existing risk warning methods consume a lot of manpower, and the efficiency of manual extraction is very low. There are large differences in experience in manual judgment. It is impossible to capture the interaction effects between risk characteristics and to prospectively explore related risks. Therefore, the existing credit risk warning methods cannot meet the actual warning needs of financial institutions. Summary of the invention

[0005] The present invention provides a credit risk early warning method to achieve accurate early warning of credit risks of financial institutions.

[0006] According to a first aspect of the present invention, there is provided a credit risk early warning method, comprising: obtaining credit-related information, wherein the credit-related information includes credit file data, customer loan repayment business information and customer-related information network;

[0007] Identify the credit file data through the information recognition model to obtain structured key element information;

[0008] Determine risk warning customers based on the key element information and the customer loan repayment business process information;

[0009] Potential associated risk customers are predicted based on the risk warning customers and the customer associated information network, and credit risk warnings are issued based on the risk warning customers and the associated risk customers.

[0010] Optionally, obtain credit-related information, including:

[0011] Extracting text materials, video resources and picture materials accumulated during the credit business processing from the local database, and using the text materials, video materials and picture materials as the credit file materials;

[0012] Acquire account-related change information involved in the customer's loan repayment business process, and use the account-related change information as the customer's loan repayment business information;

[0013] Extracting the customer association information network from a local database, wherein the customer association information network includes association relationships with each customer based on equity relationships, guarantee relationships, security relationships, and funding relationships;

[0014] The credit file data, the customer loan repayment business information and the customer's internal and external related information network are used as the credit related information.

[0015] Optionally, the information recognition model includes a combined model of optical character recognition (OCR) and natural language processing (NLP); the step of identifying the credit file data through the information recognition model to obtain structured key element information includes:

[0016] Recognize the credit file data through the OCR and NLP combined model to obtain structured recognition text;

[0017] The key element information is obtained by performing natural language processing on the structured recognized text through the OCR and NLP combined model.

[0018] Optionally, determining risk warning customers according to the key element information and the customer loan repayment business process information includes:

[0019] Inputting the key element information and the customer business process information into an expert rule model library;

[0020] Identifying target key element information and target customer business process information associated with each customer from the key element information and the customer business process information through the expert rule model library;

[0021] The risk warning customer is determined based on the target key element information and the target customer business process information.

[0022] According to another aspect of the present invention, a credit risk early warning device is provided, comprising: a credit-related information acquisition module, used to acquire credit-related information, wherein the credit-related information includes credit file data, customer loan repayment business information and customer-related information network;

[0023] A key element information identification module, used to identify the credit file data through an information identification model to obtain structured key element information;

[0024] A risk warning customer determination module is used to determine risk warning customers based on the key element information and the customer loan repayment business process information;

[0025] The credit risk warning module is used to predict potential associated risk customers based on the risk warning customers and the customer associated information network, and to issue credit risk warnings based on the risk warning customers and the associated risk customers.

[0026] According to another aspect of the present invention, there is provided an electronic device, the electronic device comprising:

[0027] at least one processor; and

[0028] a memory communicatively connected to the at least one processor; wherein,

[0029] The memory stores a computer program that can be executed by the at least one processor. The computer program is executed by the at least one processor so that the at least one processor can perform the method described in any embodiment of the present invention.

[0030] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method described in any embodiment of the present invention when executed.

[0031] According to another aspect of the present invention, a computer program product is provided, comprising a computer program, and when the computer program is executed by a processor, the method described in any embodiment of the present invention is implemented.

[0032] The beneficial technical effect of the present invention is that structured key element information is automatically identified from credit file materials through an information recognition model, thereby saving labor costs and improving recognition accuracy, and it can predict potential associated risk customers based on the acquired risk warning customers and the customer related information network, thereby achieving forward-looking risk warning, which is conducive to business personnel taking effective measures to prevent or resolve risks in a timely manner, and helping to ensure the safety of credit assets of financial institutions.

[0033] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0035] Figure 1 is a flow chart of a credit risk early warning method provided according to Embodiment 1 of the present invention;

[0036] Figure 2 is a flow chart of a credit risk early warning method provided according to Embodiment 2 of the present invention;

[0037] Figure 3 This is a schematic diagram of the structure of a credit risk early warning device provided according to Embodiment 3 of the present invention;

[0038] Figure 4 It is a schematic diagram of the structure of an electronic device provided by Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0040] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. In addition, the collected information is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data are all in compliance with the relevant laws, regulations and standards of the relevant countries and regions, necessary confidentiality measures are taken, and public order and good customs are not violated, and corresponding operation entrances are provided for users to choose to authorize or refuse

[0041] Embodiment 1

[0042] Figure 1 A flowchart of a data desensitization method is provided for the first embodiment of the present invention. This embodiment is applicable to the case of desensitizing data. The method can be executed by a data desensitization device, which can be implemented in the form of hardware and / or software. Figure 1 As shown, the method includes:

[0043] Step S101, obtaining credit association information.

[0044] Optionally, obtaining credit-related information includes: extracting text materials, video resources and picture materials accumulated when handling credit business from a local database, and using the text materials, video materials and picture materials as credit file materials; obtaining account-related change information involved in the customer's loan repayment business process, and using the account-related change information as the customer's loan repayment business information; extracting a customer-related information network from a local database, wherein the customer-related information network includes relationships with each customer based on equity relationships, guarantee relationships, security relationships and financial relationships; using credit file materials, customer loan repayment business information and the customer's internal and external related information network as credit-related information.

[0045] Specifically, in this implementation, all credit file information will be extracted from the local database, and the credit file information mainly records the textual information, video information and picture information accumulated by the customer when handling the credit business. For example, the textual information includes the loan contract and the customer notice of the loan project, etc.; the video information can be the audio and video recording when handling the credit business, and the picture information can be the identity verification certificate photo of the customer when handling the credit business, etc. Of course, this implementation is only an example, and does not limit the specific content of the credit file information. As long as it is related to the customer's credit business, it is within the scope of protection of this application. In addition, in this implementation, the account-related change information involved in the customer's loan repayment business process will also be obtained, and the account-related change information will be used as the customer's loan repayment business information, such as the monthly outflow amount and the monthly inflow amount of the account. In this embodiment, a customer association information network is also constructed based on the customer's equity relationship, guarantee relationship, security relationship and capital relationship, wherein each point in the customer association information network represents a customer entity, and customers with association relationships are connected by straight lines, and the connecting line represents the specific type of association between two customers, and the closeness of such relationships between adjacent customers is identified by a closeness value. For example, customer A and customer B are connected, and the guarantee relationship between customer A and customer B means that when customer A takes out a loan, customer B guarantees customer A's local loan. When customer A is unable to afford the loan amount, customer B's house will be used as collateral for guarantee, thereby ensuring the asset security of the financial institution. In addition, the customer association information network includes all loan customers who handle credit business in the financial institution, as well as non-loan customers who are associated with loan customers but have not handled credit business in the financial institution. In this embodiment, the number of nodes, i.e., customers, included in the customer association information network is not limited.

[0046] Step S102: Identify the credit file data through the information identification model to obtain structured key element information.

[0047] Optionally, the information recognition model includes an optical character recognition (OCR) and natural language processing (NLP) combined model; the credit file data is recognized through the information recognition model to obtain structured key element information, including: recognizing the credit file data through the OCR and NLP combined model to obtain structured recognized text; performing natural language processing on the structured recognized text through the OCR and NLP combined model to obtain key element information.

[0048] Specifically, OCR refers to the process of automatically identifying documents, vouchers, bills and other image information by computer systems and converting them into structured data that can be directly processed by computers; NLP is an important direction in the fields of computer science and artificial intelligence. It mainly studies various theories and methods that can achieve effective communication between people and computers using natural language. It is currently widely used in information extraction, text mining, information retrieval, natural language generation, dialogue systems, etc. In this implementation, the two technologies are combined to construct an OCR and NLP combined model to achieve key content recognition in credit file materials and structured extraction of business element information.

[0049] Among them, in this implementation, samples will be used to train the initial OCR and NLP combination model, and for elements with low recognition accuracy, the manually corrected structure will be fed back to the OCR and NLP combination model by supplementing annotated samples, or with the help of the correction function provided by the system platform, thereby ensuring continuous iterative optimization of the OCR and NLP combination model.

[0050] In a specific implementation, after the pre-training of the OCR and NLP combined model is completed, the credit file data will be input into the OCR and NLP combined model, and the credit file data will be identified within the model to obtain structured recognition text, and then natural language processing will be performed based on the recognition text to obtain key element information, for example, the monthly outflow amount of the account is 3,000, and the monthly outflow amount of the account is 2,000. Of course, this embodiment is only an example, and the specific content of the key element information is not limited. In addition, in this embodiment, the key element information in the credit file data is automatically extracted by the model, and compared with the manual extraction method, both the recognition accuracy and the recognition efficiency are significantly improved.

[0051] Step S103, determining risk warning customers based on key factor information and customer loan repayment business process information.

[0052] Optionally, risk warning customers are determined based on key factor information and customer loan repayment business process information, including: inputting the key factor information and customer business process information into an expert rule model library; identifying target key factor information and target customer business process information associated with each customer from the key factor information and customer business process information through the expert rule model library; and determining risk warning customers based on the target key factor information and target customer business process information.

[0053] Specifically, in this implementation, after identifying structured key element information from the credit file data, all key element information and customer loan repayment business process information obtained will be input into the expert rule model library, and the target key element information and target customer business process information associated with each customer will be identified from the expert rule model library. Since the key element information reflects the approval of the business and pre-loan agreements of the financial institution at the beginning of the credit business, the requirements for abnormal monitoring of accounts during the term, etc., and the customer business process information reflects the objective situation of customer repayments and customer manager performance during the actual business process, after obtaining the target key element information and target customer business process information associated with each customer, risk warning customers can be determined based on the target key element information and target customer business process information. For example, in this implementation, when the target key element associated with customer X is A and the target customer business process information is B, a comparison is made based on A and B to determine whether customer X is a risk warning customer. Of course, in this implementation, only one set of target key elements and target customer business process information is used as an example for explanation. In actual applications, multiple sets of target key elements and target customer business process information are usually involved for each customer. Therefore, it is necessary to determine whether customer X is a risk warning customer based on the comparison results of multiple sets of target key elements and target customer business process information.

[0054] In a specific implementation, when only a set of target key elements A and target customer business process information B are obtained for customer X, the dynamic monitoring threshold u corresponding to the key element A is also obtained from the expert rule model library, and the business monitoring value is extracted from the target customer business process information B, and the business monitoring value extracted from the target customer business process information B is compared with the target key element A to obtain the comparison difference v, and it is determined whether the comparison difference v is greater than the dynamic monitoring threshold u. If so, customer X is determined to be a risk warning customer, otherwise, customer X is determined not to be a risk warning customer.

[0055] In another specific implementation, when two sets of target key elements A1 and target customer business process information B1, target key elements A2 and target customer business process information B2 are obtained for customer X, and the dynamic monitoring threshold u1 corresponding to the key element A1 and the dynamic monitoring threshold u2 corresponding to the key element A2 are obtained from the expert rule model library, in addition, the business monitoring value is extracted from the target customer business process information B1, and the business monitoring value is extracted from the target customer business process information B2, and then the business monitoring value extracted from the target customer business process information B1 is compared with the target key element A1 to obtain the comparison difference. Value v1, compare the business monitoring value extracted from the target customer business process information B2 with the target key element A2 to obtain the comparison difference v2. When it is determined that the comparison difference v1 is greater than the dynamic monitoring threshold u1, or the comparison difference v2 is greater than the dynamic monitoring threshold u2, it is determined that customer X is a risk warning customer. Only when both comparison differences are less than the corresponding dynamic monitoring thresholds, it is determined that customer X is not a risk warning customer. Of course, in this implementation, only customer X is used as an example for explanation. The method for determining whether other customers are risk warning customers is roughly the same, and will not be repeated in this implementation.

[0056] Optionally, risk warning customers are determined based on target key element information and target customer business process information, including: obtaining a dynamic monitoring threshold corresponding to the target key element information, and extracting a business monitoring value from the target customer business process information; comparing the target key element information with the business monitoring value to obtain a comparison difference; when the comparison difference is greater than the dynamic monitoring threshold, the customer is determined to be a risk warning customer.

[0057] Among them, the dynamic monitoring threshold in this implementation mode can change with the latest requirements of the current market and industry. Because the credit project is closely related to the market and industry development, the use of a fixed monitoring threshold throughout the project life cycle will lead to errors in the monitoring results. Therefore, a dynamic monitoring threshold is used in this implementation mode, that is, firstly, a specified algorithm is used to determine the monitoring threshold in the initial state according to the credit business parameters, and then a human-computer interaction interface is provided to facilitate the adjustment of the monitoring threshold on the human-computer interaction interface according to expert experience and the current market actual situation, so that the risk warning customers can change at any time with the actual market situation, which is more in line with the current risk determination needs. For example, customer X is determined not to be a risk warning customer this month, but in the next month, due to market changes, such as an increase in loan interest rates, when customer A is re-determined, he may be determined as a risk customer. Of course, this implementation mode is only an example, and does not limit the dynamic adjustment method of the monitoring threshold.

[0058] It should be noted that since there are a large number of customers who handle credit business in financial institutions, in this embodiment, after judging each customer in the above manner, the number of risk warning customers obtained may be multiple. Therefore, in this embodiment, a risk warning customer list will be generated after all customers are judged.

[0059] Step S104, predicting potential associated risk customers based on the risk warning customers and the customer associated information network, and issuing credit risk warnings based on the risk warning customers and the associated risk customers.

[0060] Optionally, predicting potential associated risk customers based on risk warning customers and the customer associated information network includes: obtaining a pre-specified query range for the customer associated information network; searching for associated customers who have an associated relationship with the risk customers and intimacy values ​​with the associated customers within the query range; and treating associated customers whose intimacy values ​​are greater than a specified value as associated risk customers.

[0061] Specifically, in this embodiment, not only a group of risk warning customers will be directly determined based on the identified key element information and loan repayment business process information, but also potential associated risk customers will be predicted from the customer associated information network based on the determined risk warning customers. Since the associated information network is very large, each customer has a direct or indirect association relationship with the other customers in the associated information network. Therefore, in this embodiment, the query range can be pre-specified according to the prediction accuracy, for example, specifying to mine associations within 4 degrees, where, when the connection relationship between customers is determined to be: customer M-customer N-customer P-customer Q-customer X, it is determined that the distance between customer M and customer N is 1 degree, the distance between customer M and customer P is 2 degrees, the distance between customer M and customer Q is 3 degrees, and the distance between customer M and customer X is 4 degrees. Since the closeness between adjacent customers is identified by the intimacy value between customers, in this embodiment, for each risk warning customer in the risk warning customer list, a search will be performed within the specified query range in the customer associated information network to obtain customers who have an association relationship with each risk warning customer and whose intimacy value is greater than 9, and the searched customers are used as associated risk customers.

[0062] It should be noted that, in this embodiment, after searching all the customers in the risk warning customer list, all the associated risk customers will be obtained, and a potential associated risk customer list will be generated based on all the associated risk customers obtained. Among them, the potential associated risk customer list is a situation where there is no credit risk at present, but there may be credit risk in the future. For example, when customer M is determined to be a directly determined risk warning customer, customer N is a potential associated risk customer obtained through search based on customer M, and customer M and customer N are in a guarantee relationship. When customer M becomes a dishonest customer due to insufficient credit repayment ability, customer N's ​​property needs to be mortgaged, thereby causing asset loss of customer N. When customer N is also a credit customer of a financial institution, when the asset loss occurs, it is very likely to affect customer N's ​​own credit business, thereby causing customer N's ​​repayment ability to decline, making customer N a dishonest customer. In this embodiment, these customers can be mined out first, thereby further ensuring the asset security of financial institutions.

[0063] It is worth mentioning that in this implementation, after risk warning customers and potential associated risk customers are determined based on the existing credit-related information, a credit safety alarm prompt will be generated. The credit safety alarm prompt identifies customers with credit safety risks, thereby prompting business processing personnel to pay special attention to the identified customers when handling business.

[0064] In the implementation mode of the present application, structured key element information is automatically identified from credit file materials through an information recognition model, thereby saving labor costs and improving recognition accuracy. It is also possible to predict potential associated risk customers based on the acquired risk warning customers and the customer related information network, thereby achieving forward-looking risk warning, which is conducive to business personnel taking effective measures to prevent or resolve risks in a timely manner and helping to ensure the safety of credit assets of financial institutions.

[0065] Embodiment 2

[0066] Figure 2 This is a flowchart of a credit risk warning method provided by Embodiment 2 of the present invention. This embodiment is based on the above embodiment. After predicting potential associated risk customers based on risk warning customers and customer associated information network, it also includes: setting a designated verification method for associated risk customers; when receiving a business application from an associated risk customer, performing business security verification on the associated risk customer based on the designated verification method. Figure 2 As shown, the method includes:

[0067] Step S201, obtaining credit association information.

[0068] Optionally, obtaining credit-related information includes: extracting text materials, video resources and picture materials accumulated when handling credit business from a local database, and using the text materials, video materials and picture materials as credit file materials; obtaining account-related change information involved in the customer's loan repayment business process, and using the account-related change information as the customer's loan repayment business information; extracting a customer-related information network from a local database, wherein the customer-related information network includes relationships with each customer based on equity relationships, guarantee relationships, security relationships and financial relationships; using credit file materials, customer loan repayment business information and the customer's internal and external related information network as credit-related information.

[0069] Step S202: Identify the credit file data through the information identification model to obtain structured key element information.

[0070] Optionally, the information recognition model includes an optical character recognition (OCR) and natural language processing (NLP) combined model; the credit file data is recognized through the information recognition model to obtain structured key element information, including: recognizing the credit file data through the OCR and NLP combined model to obtain structured recognized text; performing natural language processing on the structured recognized text through the OCR and NLP combined model to obtain key element information.

[0071] Step S203, determining risk warning customers based on key factor information and customer loan repayment business process information.

[0072] Optionally, risk warning customers are determined based on key factor information and customer loan repayment business process information, including: inputting the key factor information and customer business process information into an expert rule model library; identifying target key factor information and target customer business process information associated with each customer from the key factor information and customer business process information through the expert rule model library; and determining risk warning customers based on the target key factor information and target customer business process information.

[0073] Optionally, risk warning customers are determined based on target key element information and target customer business process information, including: obtaining a dynamic monitoring threshold corresponding to the target key element information, and extracting a business monitoring value from the target customer business process information; comparing the target key element information with the business monitoring value to obtain a comparison difference; when the comparison difference is greater than the dynamic monitoring threshold, the customer is determined to be a risk warning customer.

[0074] Step S204, predicting potential associated risk customers based on the risk warning customers and the customer associated information network, and issuing credit risk warnings based on the risk warning customers and the associated risk customers.

[0075] Optionally, predicting potential associated risk customers based on risk warning customers and the customer associated information network includes: obtaining a pre-specified query range for the customer associated information network; searching for associated customers who have an associated relationship with the risk customers and intimacy values ​​with the associated customers within the query range; and treating associated customers whose intimacy values ​​are greater than a specified value as associated risk customers.

[0076] Step S205, setting a designated verification method for the associated risk customer, and when receiving a business application from the associated risk customer, performing business security verification on the associated risk customer based on the designated verification method.

[0077] Specifically, in this implementation, after potential associated risk customers are predicted based on risk warning customers, since the associated risk customers may have insufficient loan repayment capacity, a more stringent designated verification method will be set for these customers compared to ordinary customers to ensure the asset security of the financial institution.

[0078] For example, when ordinary customers apply for initial credit, only the salary flow can be verified. However, when potential risk-associated customers apply for initial credit, not only the salary flow but also the asset mortgage status within the previous five years must be verified. Therefore, compared with ordinary customers, the verification method for potential risk-associated customers is more stringent and detailed. Therefore, when receiving a business application from a risk-associated customer, the customer will be subject to business security verification based on the specified verification method set above.

[0079] It should be noted that, in the present embodiment, when an associated risk customer is verified according to a specified verification method and the verification is passed, it means that the customer does not have a credit risk. At this time, the customer will be deleted from the list of potential associated risk customers, and when the customer subsequently executes the loan repayment business process, the customer will also be monitored as an ordinary safe customer, thereby reducing the use of memory space and excessive occupation of computing resources; in addition, when an associated risk customer is verified according to a specified verification method and the verification is not passed, it means that the customer is unable to repay the credit. At this time, a credit security alarm will be generated to prompt the business processing personnel to suspend the credit business processing for the customer.

[0080] In the implementation mode of the present application, structured key element information is automatically identified from credit file materials through an information recognition model, thereby saving labor costs and improving recognition accuracy. It is also possible to predict potential associated risk customers based on the acquired risk warning customers and the customer related information network, thereby achieving forward-looking risk warning, which is conducive to business personnel taking effective measures to prevent or resolve risks in a timely manner and helping to ensure the safety of credit assets of financial institutions.

[0081] Embodiment 3

[0082] Figure 3 This is a schematic diagram of the structure of a credit risk early warning device provided by Embodiment 3 of the present invention. Figure 3 As shown, the device includes: a credit-related information acquisition module 310 , a key element information identification module 320 , a risk warning customer determination module 330 and a credit risk warning module 340 .

[0083] The credit-related information acquisition module 310 is used to acquire credit-related information, wherein the credit-related information includes credit file data, customer loan repayment business information and customer-related information network;

[0084] The key element information identification module 320 is used to identify the credit file data through the information identification model to obtain structured key element information;

[0085] The risk warning customer determination module 330 is used to determine the risk warning customer according to the key element information and the customer loan repayment business process information;

[0086] The credit risk warning module 340 is used to predict potential associated risk customers based on risk warning customers and customer associated information network, and to issue credit risk warnings based on risk warning customers and associated risk customers.

[0087] Optionally, a credit-related information acquisition module is used to extract text materials, video resources and picture materials accumulated during the credit business processing from the local database, and use the text materials, video materials and picture materials as credit file materials;

[0088] Obtain the account-related change information involved in the customer's loan repayment business process, and use the account-related change information as the customer's loan repayment business information;

[0089] Extracting a customer-related information network from a local database, wherein the customer-related information network includes a relationship with each customer based on equity relationship, guarantee relationship, security relationship, and capital relationship;

[0090] Credit file data, customer loan repayment business information and the customer's internal and external related information network are used as credit related information.

[0091] Optionally, the information recognition model includes an optical character recognition (OCR) and natural language processing (NLP) combined model; a key element information recognition module is used to recognize credit file data through the OCR and NLP combined model to obtain structured recognition text;

[0092] The key element information is obtained by performing natural language processing on the structured recognized text through the OCR and NLP combined model.

[0093] Optionally, a risk warning customer determination module is used to input key element information and customer business process information into the expert rule model library;

[0094] Identify target key element information and target customer business process information associated with each customer from key element information and customer business process information through the expert rule model library;

[0095] Identify risk warning customers based on target key element information and target customer business process information.

[0096] Optionally, the risk warning customer determination module is further used to obtain a dynamic monitoring threshold corresponding to the target key element information and extract a business monitoring value from the target customer business process information;

[0097] Compare the target key element information with the business monitoring value to obtain the comparison difference;

[0098] When the comparison difference is greater than the dynamic monitoring threshold, the customer is determined to be a risk warning customer.

[0099] Optionally, a credit risk early warning module is used to obtain a pre-specified query scope for the customer-related information network;

[0100] Search for associated customers who are associated with risky customers within the query scope, as well as the intimacy values ​​with the associated customers;

[0101] Associated customers whose intimacy values ​​are greater than a specified value are considered associated risk customers.

[0102] Optionally, the device further comprises a security verification module for setting a specified verification method for associated risk customers;

[0103] When receiving a business application from an associated risk customer, a business security verification is performed on the associated risk customer based on a specified verification method.

[0104] The credit risk early warning device provided in the embodiment of the present invention can execute the credit risk early warning method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0105] Embodiment 4

[0106] Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0107] like Figure 4 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0108] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0109] The processor 11 may be any general and / or dedicated processing component with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the credit risk early warning method.

[0110] In some embodiments, the credit risk early warning method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the credit risk early warning method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the credit risk early warning method in any other appropriate manner (e.g., by means of firmware).

[0111] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0112] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0113] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0114] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0115] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0116] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0117] Embodiment 5

[0118] An embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the credit risk early warning method provided in any embodiment of the present application.

[0119] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0120] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0121] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A credit risk early warning method, characterized in that: include: Acquiring credit-related information, wherein the credit-related information includes credit file data, customer loan repayment business information, and customer-related information network; Identify the credit file data through the information recognition model to obtain structured key element information; Determine risk warning customers based on the key element information and the customer loan repayment business process information; Potential associated risk customers are predicted based on the risk warning customers and the customer associated information network, and credit risk warnings are issued based on the risk warning customers and the associated risk customers.

2. The method according to claim 1, characterized in that The obtaining of credit-related information includes: Extracting text materials, video resources and picture materials accumulated during the credit business processing from the local database, and using the text materials, video materials and picture materials as the credit file materials; Acquire account-related change information involved in the customer's loan repayment business process, and use the account-related change information as the customer's loan repayment business information; Extracting the customer association information network from a local database, wherein the customer association information network includes association relationships with each customer based on equity relationships, guarantee relationships, security relationships, and funding relationships; The credit file data, the customer loan repayment business information and the customer's internal and external related information network are used as the credit related information.

3. The method according to claim 1, characterized in that The information recognition model includes an optical character recognition (OCR) and natural language processing (NLP) combined model; The method of identifying the credit file data through the information recognition model to obtain structured key element information includes: Recognize the credit file data through the OCR and NLP combined model to obtain structured recognition text; The key element information is obtained by performing natural language processing on the structured recognized text through the OCR and NLP combined model.

4. The method according to claim 1, characterized in that: The step of determining the risk warning customer based on the key element information and the customer loan repayment business process information includes: Inputting the key element information and the customer business process information into an expert rule model library; Identifying target key element information and target customer business process information associated with each customer from the key element information and the customer business process information through the expert rule model library; The risk warning customer is determined based on the target key element information and the target customer business process information.

5. The method according to claim 4, characterized in that The step of determining the risk warning customer according to the target key element information and the target customer business process information includes: Acquire a dynamic monitoring threshold corresponding to the target key element information, and extract a business monitoring value from the target customer business process information; Comparing the target key element information with the business monitoring value to obtain a comparison difference; When the comparison difference is greater than the dynamic monitoring threshold, the customer is determined to be the risk warning customer.

6. The method according to claim 1, characterized in that The predicting of potential associated risk customers based on the risk warning customers and the customer associated information network includes: Obtaining a pre-specified query scope for the customer-related information network; Searching for associated customers who have an associated relationship with the risky customer and intimacy values ​​with the associated customers within the query scope; The associated customers whose intimacy values ​​are greater than a specified value are regarded as the associated risk customers.

7. The method according to any one of claims 1 to 6, characterized in that: After predicting the potential associated risk customers according to the risk warning customers and the customer associated information network, the method further includes: Setting a designated verification method for the associated risk customers; When a business application of the associated risk customer is received, a business security verification is performed on the associated risk customer based on the specified verification method.

8. A credit risk early warning device, characterized in that: The device comprises: A credit-related information acquisition module, used to acquire credit-related information, wherein the credit-related information includes credit file data, customer loan repayment business information and customer-related information network; A key element information identification module, used to identify the credit file data through an information identification model to obtain structured key element information; A risk warning customer determination module is used to determine risk warning customers based on the key element information and the customer loan repayment business process information; The credit risk warning module is used to predict potential associated risk customers based on the risk warning customers and the customer associated information network, and to issue credit risk warnings based on the risk warning customers and the associated risk customers.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method according to any one of claims 1 to 7 when executed.

11. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 7 when being executed by a processor.

Images