Credit risk assessment method and device based on artificial intelligence, equipment and medium

Abstract

The invention relates to the technical field of financial information, in particular to an artificial intelligence-based credit risk assessment method, device and equipment and a medium, and the artificial intelligence-based credit risk assessment method comprises the following steps: obtaining a credit application request triggered by a user, obtaining user basic information from the credit application request, extracting user active data from the user basic data; obtaining a current position of a user from the credit application request, and performing calculation according to the current position of the user to obtain passive data of the user; generating and triggering a user interaction verification request to a user, and obtaining corresponding user interaction data according to the user interaction verification request; and performing calculation according to the user active data, the user passive data and the user interaction data to obtain a user risk assessment score, and generating a user credit risk control result according to the user risk assessment score. The method has the effect of improving the safety and convenience of credit risk assessment and approval.

Description

Technical Field

[0001] The present application relates to the technical field of financial information, and in particular to a credit risk assessment method, device, equipment and medium based on artificial intelligence. Background Art

[0002] At present, with the global financial industry accelerating its digital transformation, technological innovation in the field of credit services is advancing at an unprecedented speed. The application of modern technologies, such as big data analysis, artificial intelligence and biometrics, is reshaping the financial services landscape, improving operational efficiency and improving customer experience. However, despite the significant progress made in financial technology, there are still many challenges in the credit approval process, especially in balancing approval speed and risk control.

[0003] At present, many financial institutions use traditional risk assessment methods based on rule engines, supplemented by basic artificial intelligence algorithms, to improve the efficiency of credit decision-making. Such solutions usually involve basic personal financial data review, such as credit history and income level, as the main basis for judging the borrower's willingness and ability to repay.

[0004] The above-mentioned prior art solutions have the following defects: This traditional assessment method that relies too much on structured data ignores the value of unstructured data, such as the user's digital footprint and social network connectivity. In addition, static risk assessment models are difficult to respond to market changes and the evolution of personal behavior patterns in a timely manner. Especially in the era of mobile Internet and the Internet of Things, diversified data such as users' location information and consumption habits are crucial for a comprehensive assessment of their credit status. Furthermore, the single-level identity authentication mechanism is vulnerable to fraud attacks and cannot meet highly personalized customer needs, affecting the security and convenience of the entire approval process. Summary of the invention

[0005] In order to improve the security and convenience of credit risk assessment and approval, the present application provides a credit risk assessment method, device, equipment and medium based on artificial intelligence.

[0006] The above-mentioned invention objective of the present application is achieved through the following technical solutions: A credit risk assessment method based on artificial intelligence, the credit risk assessment method based on artificial intelligence comprising: Obtaining a credit application request triggered by a user, obtaining basic user information from the credit application request, and extracting active user data from the basic user data; Obtaining the user's current location from the credit application request, and performing calculations based on the user's current location to obtain user passive data; Generate and trigger a user interaction verification request to the user, and obtain corresponding user interaction data according to the user interaction verification request; Calculation is performed based on the user active data, user passive data and user interactive data to obtain a user risk assessment score, and a user credit risk control result is generated based on the user risk assessment score.

[0007] By adopting the above technical solution, accurate assessment and efficient approval of credit risk are achieved. First, in the process of obtaining the credit application request triggered by the user, the basic information of the user can be automatically extracted from the credit application request, including but not limited to name, ID number, contact information, etc. Further, the data actively provided by the user can be extracted from the user's basic data, so that the basic situation of the user can be grasped at the first time, laying a solid data foundation for subsequent risk assessment. Secondly, by obtaining and calculating the user's current location, the user's passive data is obtained, and the user's current location is compared with the preset high-risk area, and the distance between the user and these high-risk areas is calculated, and the risk level is adjusted accordingly. The location-based risk assessment method can more accurately reflect the impact of the user's environment on his credit status, thereby improving the accuracy of risk assessment. Thirdly, a user interaction verification request is generated and triggered to the user, and the user's integrity and cooperation can be calculated. Through a multi-level identity authentication mechanism, identity impersonation and fraud can be effectively prevented, and security is enhanced. Finally, the user's active data, user passive data and user interaction data are comprehensively calculated to obtain the user's risk assessment score. This process involves the setting of multiple weight parameters, including active weight, passive weight and interactive weight. The system will perform a weighted summation of various data according to the pre-set weight distribution, and finally obtain the user's risk assessment score. Based on this score, the corresponding credit risk control results can be generated, which not only takes into account the user's economic situation, but also takes into account the user's behavioral habits and environmental factors, making the risk assessment more comprehensive and accurate.

[0008] In a preferred example, the present application may be further configured as follows: the current location of the user is obtained from the credit application request, and the user passive data is obtained by calculation based on the current location of the user, specifically including: Obtaining preset high-risk areas, and obtaining the center position of each of the high-risk areas as the corresponding high-risk position; The risk distance between the user's current location and each of the high-risk locations is obtained, and the risk distance with the smallest value is input into the following formula to calculate the user's passive data: , where R s refers to the user passive data, d refers to the minimum risk distance, σ refers to the preset risk curve coefficient, and k refers to the preset risk coefficient.

[0009] By adopting the above technical solutions, more accurate user risk assessment can be achieved, especially for the potential risks brought by the user's geographical location. First, by obtaining the user's current location and comparing it with the pre-set high-risk areas, it is possible to effectively identify whether the user is in or close to these high-risk areas. High-risk areas are usually determined based on historical data and expert knowledge bases, such as areas with high crime rates, areas with frequent natural disasters, or areas with unstable economies. This not only takes into account the user's current specific location, but also combines the historical risk data of these areas, making the risk assessment more comprehensive and accurate. Secondly, by calculating the distance between the user's current location and the center of each high-risk area, the granularity of the risk assessment is further refined, which can more carefully reflect the actual risk level of the user's environment. For example, even if the user is not within the geographical boundary of a high-risk area, if he is very close to the area, he may still be affected by adverse events in the area. Therefore, this method can capture potential risk points that traditional methods may overlook. Again, the minimum risk distance is substituted into a specific formula for calculation, and the user passive data is obtained. The settings of parameters k and σ can be adjusted according to actual business needs to adapt to different types of high-risk areas and their risk characteristics, thereby maintaining the rationality of risk assessment. In summary, it is not only possible to more accurately assess the potential risks of users due to their geographical locations during the credit approval process, but also to flexibly adjust risk assessment according to different business needs, thereby improving the scientificity and reliability of credit approval.

[0010] In a preferred example, the present application may be further configured as follows: generating and triggering a user interaction verification request to a user, and obtaining corresponding user interaction data according to the user interaction verification request, specifically including: The user interaction verification request includes a biometric verification request and a question-and-answer verification request, and the biometric verification feature and the question-and-answer verification reply data are obtained from the biometric verification request and the question-and-answer verification request respectively; The biometric verification feature and the question-and-answer verification response data are calculated respectively to obtain the user interaction data.

[0011] By adopting the above technical solution, multi-level verification of user identity can be achieved, thereby greatly improving the accuracy and security of credit risk assessment. Specifically: when a user triggers a credit application request, an interactive verification request including biometric verification will be automatically sent to the user. The user needs to complete the verification process through fingerprints, facial recognition or other biometrics, and then analyze and compare these features through a specific algorithm to obtain a total biometric verification score, which can reflect the degree of match between the biometrics provided by the user and the standard features stored in the system, thereby effectively preventing identity fraud and fraud. In addition to biometric verification, a series of pre-set questions are sent to the user and analyzed based on the user's answers, which can not only verify the user's identity, but also further confirm the user's understanding of their own information, thereby improving the accuracy of risk assessment. By adopting a multi-level identity authentication mechanism, this technical solution can not only effectively prevent identity fraud and fraud, but also improve the accuracy and reliability of credit risk assessment, thereby providing financial institutions with more efficient and secure credit approval services. This not only helps to improve the operational efficiency of financial institutions, but also enhances user experience and improves customer satisfaction and loyalty.

[0012] In a preferred example, the present application may be further configured as follows: the biometric verification feature and the question-and-answer verification response data are calculated respectively to obtain the user interaction data, specifically including: The biometric verification feature is input into the following formula to obtain the biometric verification score: , where C is the biometric verification score, n is the number of user feature types taken, s is the corresponding feature similarity, and w is the weight coefficient corresponding to the feature; The question-and-answer verification response data is input into the following formula to obtain the question-and-answer verification score: , where A refers to the question-answering verification score of selecting the template with the minimum edit distance ED given a user question Q and a set of predefined answer templates (T={t_1, t_2, ..., t_n}); The user interaction data is generated according to the biometric verification score and the question-and-answer verification score.

[0013] By adopting the above technical solution, the biometric verification features and question-answer verification response data are accurately calculated, and more reliable user interaction data is obtained. First, for the biometric verification features, by inputting them into a specific formula, the similarity score of each feature can be effectively quantified, and the corresponding weight coefficient can be assigned, and finally a comprehensive biometric verification score is calculated, which can not only fully reflect the identity characteristics of the user, but also prevent identity impersonation and fraud to a certain extent. Secondly, when processing the question-answer verification response data, by comparing the edit distance between the user's question and the predefined answer template, the template with the smallest edit distance is selected as the best matching answer, ensuring the accurate understanding and response of the user's question. This method based on natural language processing technology greatly improves the answer accuracy of intelligent customer service and enhances the user experience. Finally, the biometric verification score and the question-answer verification score are organically combined to form complete user interaction data, which provides strong support for subsequent risk assessment. In summary, this technical solution realizes the effective capture and accurate assessment of user interaction data through a scientific and reasonable calculation method, and further improves the accuracy and reliability of credit risk assessment.

[0014] In a preferred example, the present application can be further configured as follows: the calculation is performed based on the user active data, the user passive data and the user interactive data to obtain a user risk assessment score, and the user credit risk control result is generated based on the user risk assessment score, specifically including: Obtaining active weights, passive weights and interactive weights corresponding to the user active data, the user passive data and the user interactive data; The user active data, the active weight, the user passive data, the passive weight, the user interactive data and the interactive weight are weighted to obtain the user risk assessment score.

[0015] By adopting the above technical solution, the user's active data, passive data and interactive data are weighted to form a user risk assessment score, making the credit risk assessment more comprehensive and accurate. Specifically, the user's active data covers the financial information and asset proof provided by the user, which directly reflects their economic status, while the passive data includes the user's geographical location, social activities and other data that indirectly reflect their behavior patterns, and the interactive data is generated through real-time interaction with the user. These data together constitute a multi-dimensional evaluation system. This method can not only more accurately judge the user's repayment ability and willingness, but also effectively identify potential fraudulent behavior, thereby improving the accuracy of credit approval. Secondly, by using the weighted calculation method to generate the user risk assessment score, different weights can be assigned according to the importance and reliability of different types of data, thereby ensuring the scientificity and rationality of the final evaluation results. By flexibly adjusting the weights of various data, it can better adapt to the risk assessment needs in different scenarios, enhancing flexibility and adaptability. Thirdly, the user credit risk control results are generated according to the user risk assessment score, realizing the automation and intelligence of the credit approval process. The traditional credit approval process often requires a lot of manual participation, which is time-consuming and easily affected by subjective factors. This technical solution, through automated processing and intelligent analysis, can efficiently process a large number of applications in a short period of time, greatly shortening the approval time and improving work efficiency. At the same time, due to the reduction of human intervention, the probability of operational errors is also reduced, further improving the fairness and transparency of the approval process.

[0016] The second object of the invention is achieved by the following technical solutions: A credit risk assessment device based on artificial intelligence, the credit risk assessment device based on artificial intelligence comprising: An active data acquisition module, used to acquire a credit application request triggered by a user, acquire user basic information from the credit application request, and extract user active data from the user basic data; A passive data acquisition module, used to acquire the user's current location from the credit application request, and perform calculations based on the user's current location to obtain user passive data; An interaction data acquisition module, used to generate and trigger a user interaction verification request to the user, and acquire corresponding user interaction data according to the user interaction verification request; The risk control assessment module is used to calculate based on the user's active data, user's passive data and user's interactive data to obtain a user risk assessment score, and generate a user credit risk control result based on the user's risk assessment score.

[0017] By adopting the above technical solution, accurate assessment and efficient approval of credit risk are achieved. First, in the process of obtaining the credit application request triggered by the user, the basic information of the user can be automatically extracted from the credit application request, including but not limited to name, ID number, contact information, etc. Further, the data actively provided by the user can be extracted from the user's basic data, so that the basic situation of the user can be grasped at the first time, laying a solid data foundation for subsequent risk assessment. Secondly, by obtaining and calculating the user's current location, the user's passive data is obtained, and the user's current location is compared with the preset high-risk area, and the distance between the user and these high-risk areas is calculated, and the risk level is adjusted accordingly. The location-based risk assessment method can more accurately reflect the impact of the user's environment on his credit status, thereby improving the accuracy of risk assessment. Thirdly, a user interaction verification request is generated and triggered to the user, and the user's integrity and cooperation can be calculated. Through a multi-level identity authentication mechanism, identity impersonation and fraud can be effectively prevented, and security is enhanced. Finally, the user's active data, user passive data and user interaction data are comprehensively calculated to obtain the user's risk assessment score. This process involves the setting of multiple weight parameters, including active weight, passive weight and interactive weight. The system will perform a weighted summation of various data according to the pre-set weight distribution, and finally obtain the user's risk assessment score. Based on this score, the corresponding credit risk control results can be generated, which not only takes into account the user's economic situation, but also takes into account the user's behavioral habits and environmental factors, making the risk assessment more comprehensive and accurate.

[0018] The third objective of the present application is achieved through the following technical solutions: A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned artificial intelligence-based credit risk assessment method when executing the computer program.

[0019] The fourth objective of the present application is achieved through the following technical solutions: A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned artificial intelligence-based credit risk assessment method.

[0020] In summary, the present application includes at least one of the following beneficial technical effects: 1. By obtaining active data, passive data and interactive data from users and using the weighted average method for comprehensive calculation, a user risk assessment score is formed, which enables a comprehensive assessment of the borrower's credit status and significantly improves the accuracy and efficiency of credit approval; 2. The introduction of location scene recognition function dynamically adjusts the risk level according to the distance between the user's current location and the high-risk area, which can effectively deal with the potential risks brought about by changes in the user's geographical location and enhance the flexibility and real-time nature of risk control; 3. The multi-level identity authentication mechanism combines biometric verification and question-and-answer verification. By calculating the similarity score of biometrics and the edit distance of question-and-answer verification, it improves the accuracy and security of identity authentication, while simplifying the user operation steps and improving user experience and satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flow chart of a credit risk assessment method based on artificial intelligence in one embodiment of the present application; Figure 2 It is a principle block diagram of a credit risk assessment system based on artificial intelligence in one embodiment of the present application; Figure 3 It is a schematic diagram of a device in an embodiment of the present application. DETAILED DESCRIPTION

[0022] The present application is further described in detail below in conjunction with the accompanying drawings.

[0023] In one embodiment, if Figure 1 As shown, the present application discloses a credit risk assessment method based on artificial intelligence, which specifically includes the following steps: S10: Obtain a credit application request triggered by a user, obtain basic user information from the credit application request, and extract user active data from the basic user data.

[0024] In this embodiment, the credit application request refers to a request triggered by a user to conduct a credit loan business.

[0025] Specifically, when a user wishes to apply for a credit loan, the financial institution will usually conduct a risk assessment on the user. Therefore, when the user triggers the credit application request, the credit application request is triggered by filling in the basic user information. After the credit application request is obtained, the corresponding asset data is obtained from the basic user information as the user's active data.

[0026] S20: Obtain the user's current location from the credit application request, perform calculations based on the user's current location, and obtain user passive data.

[0027] Specifically, when the user triggers the credit application request, the user's current location is obtained through the positioning function of the device that the user triggers the request, so that the credit application request carries the user's current location.

[0028] Furthermore, by obtaining the location of a preset high-risk area, the current location of the user and the location of the high-risk area are calculated to obtain user passive data.

[0029] S30: Generate and trigger a user interaction verification request to the user, and obtain corresponding user interaction data according to the user interaction verification request.

[0030] Specifically, while verifying the user, we want to trigger the verification of the corresponding biometric features of the user, such as fingerprint, face, and voiceprint, and trigger the corresponding question-and-answer interaction request, and generate the user interaction data based on the results of the user-triggered biometric verification and the results of the question-and-answer interaction.

[0031] S40: Calculate based on user active data, user passive data and user interaction data to obtain a user risk assessment score, and generate a user credit risk control result based on the user risk assessment score.

[0032] Specifically, after obtaining the user's active data, passive data and interactive data, the corresponding algorithm is used for calculation to obtain the user's risk score, and the user risk assessment score is used to generate the corresponding user credit risk control result to provide corresponding reference for financial institutions.

[0033] In this embodiment, accurate assessment and efficient approval of credit risk are achieved. First, in the process of obtaining the credit application request triggered by the user, the basic information of the user can be automatically extracted from the credit application request, including but not limited to name, ID number, contact information, etc. Further, the data actively provided by the user can be extracted from the user's basic data, so that the basic situation of the user can be grasped at the first time, and a solid data foundation can be laid for subsequent risk assessment. Secondly, by obtaining and calculating the user's current location, the user's passive data is obtained, and the user's current location is compared with the preset high-risk area, and the distance between the user and these high-risk areas is calculated, and the risk level is adjusted accordingly. The location-based risk assessment method can more accurately reflect the impact of the user's environment on his credit status, thereby improving the accuracy of risk assessment. Thirdly, a user interaction verification request is generated and triggered to the user, and the user's integrity and cooperation can be calculated. Through a multi-level identity authentication mechanism, identity impersonation and fraud can be effectively prevented, and security is enhanced. Finally, the user's active data, user passive data and user interaction data are comprehensively calculated to obtain the user's risk assessment score. This process involves the setting of multiple weight parameters, including active weight, passive weight and interactive weight. The system will perform a weighted summation of various data according to the pre-set weight distribution, and finally obtain the user's risk assessment score. Based on this score, the corresponding credit risk control results can be generated, which not only takes into account the user's economic situation, but also takes into account the user's behavioral habits and environmental factors, making the risk assessment more comprehensive and accurate.

[0034] In one embodiment, in step S20, the user's current location is obtained from the credit application request, and calculation is performed based on the user's current location to obtain the user's passive data, which specifically includes: S21: Obtain preset high-risk areas, and obtain the center position of each high-risk area as the corresponding high-risk position.

[0035] Specifically, corresponding high-risk areas are set in advance based on historical high-risk credit records and other relevant information, and the center coordinates of each high-risk area are obtained as the center position, and the center position of each high-risk area is used as the corresponding high-risk position.

[0036] S22: Obtain the risk distance between the user's current location and each high-risk location, input the risk distance with the smallest value into the following formula, and calculate the user's passive data: , where R s refers to user passive data, d refers to the minimum risk distance, σ refers to the preset risk curve coefficient, and k refers to the preset risk coefficient.

[0037] Specifically, the risk distance between the user's current location and each high-risk location is calculated, and the risk distance with the minimum value is obtained and input into the above formula to calculate the user's passive data, wherein the corresponding σ and risk coefficient k can be obtained based on the actual situation of the financial product corresponding to the credit application request and the risk level of the high-risk area corresponding to the risk distance with the minimum value, so that the calculated user passive data is more in line with the actual situation.

[0038] In one embodiment, in step S30, a user interaction verification request is generated and triggered to the user, and corresponding user interaction data is obtained according to the user interaction verification request, specifically including: S31: The user interactive verification request includes a biometric verification request and a question-and-answer verification request, and the biometric verification feature and question-and-answer verification response data are obtained from the biometric verification request and the question-and-answer verification request respectively.

[0039] Specifically, in the process of triggering a user interaction request and performing interactive verification with the user, a biometric verification request and a question-and-answer verification request are respectively triggered to the user, and the biometric verification feature and question-and-answer verification response data are obtained according to the corresponding requests.

[0040] S32: Calculate the biometric verification feature and the question-and-answer verification response data respectively to obtain user interaction data.

[0041] Specifically, the biometric verification feature and the question-and-answer verification answer are calculated using corresponding formulas, and the calculated results are used as the user interaction data.

[0042] In the embodiment, in step S32, the biometric verification feature and the question-and-answer verification response data are calculated respectively to obtain user interaction data, which specifically includes: S321: Input the biometric verification feature into the following formula to obtain a biometric verification score: , where C is the biometric verification score, n refers to the number of user feature types taken, s refers to the corresponding feature similarity, and w refers to the weight coefficient corresponding to the feature.

[0043] Specifically, for each different type of user feature, such as fingerprint features, voiceprint features and / or facial features, the corresponding feature values ​​are extracted from the corresponding features, and the similarity is compared with the pre-stored benchmark value of the user to obtain the corresponding feature similarity s, and the weight coefficient w of the different types of features is calculated to obtain the biometric verification score.

[0044] S322: Input the question and answer verification response data into the following formula to obtain the question and answer verification score: , where A refers to the question-answering verification score of selecting the template with the minimum edit distance ED given a user question Q and a set of predefined answer templates (T={t_1, t_2, ..., t_n}).

[0045] Specifically, by calculating the similarity between the verification response data corresponding to the question Q asked to the user and the answer template of the question, the one with the highest similarity is taken as the minimum edit distance ED, and then the score of each interactive question is calculated, so as to accumulate the verification score of the user.

[0046] S323: Generate user interaction data based on the biometric verification score and the question-and-answer verification score.

[0047] Specifically, the biometric verification score and the question-and-answer verification score are weighted and summed to obtain the user interaction data.

[0048] In one embodiment, in step S40, a user risk assessment score is obtained based on the user active data, the user passive data and the user interactive data, and a user credit risk control result is generated based on the user risk assessment score, specifically including: S41: Obtain active weights, passive weights and interactive weights corresponding to user active data, user passive data and user interactive data.

[0049] Specifically, according to the corresponding product in the credit application request, the security risk level of the product is obtained, and the preset active weight, passive weight and interactive weight are obtained.

[0050] S42: The user's active data, active weight, user's passive data, passive weight, user's interactive data and interactive weight are weighted to obtain a user risk assessment score.

[0051] Specifically, the user's active data, active weight, user's passive data, passive weight, user's interactive data and interactive weight are weighted and summed to obtain the user's risk assessment score.

[0052] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0053] In one embodiment, a credit risk assessment device based on artificial intelligence is provided, and the credit risk assessment device based on artificial intelligence corresponds one-to-one with the credit risk assessment method based on artificial intelligence in the above embodiment. Figure 2 As shown, the credit risk assessment device based on artificial intelligence includes an active data acquisition module, a passive data acquisition module, an interactive data acquisition module and a risk control assessment module. The detailed description of each functional module is as follows: An active data acquisition module is used to acquire a credit application request triggered by a user, obtain basic user information from the credit application request, and extract user active data from the basic user data; A passive data acquisition module is used to obtain the user's current location from the credit application request, and to perform calculations based on the user's current location to obtain the user's passive data; An interaction data acquisition module, used to generate and trigger a user interaction verification request to the user, and acquire corresponding user interaction data according to the user interaction verification request; The risk control assessment module is used to calculate based on user active data, user passive data and user interaction data to obtain user risk assessment scores, and generate user credit risk control results based on the user risk assessment scores.

[0054] Optionally, the passive data acquisition module includes: The risk location acquisition submodule is used to acquire the preset high-risk areas and obtain the center position of each high-risk area as the corresponding high-risk location; The passive data calculation submodule is used to obtain the risk distance between the user's current location and each high-risk location, and input the risk distance with the smallest value into the following formula to calculate the user's passive data: , where R s refers to user passive data, d refers to the minimum risk distance, σ refers to the preset risk curve coefficient, and k refers to the preset risk coefficient.

[0055] Optionally, the interactive data acquisition module includes: An interactive data acquisition submodule, used for user interactive verification requests including biometric verification requests and question-and-answer verification requests, and obtaining biometric verification features and question-and-answer verification response data from the biometric verification request and question-and-answer verification request respectively; The interactive data calculation submodule is used to calculate the biometric verification features and question-and-answer verification response data respectively to obtain user interaction data.

[0056] Optionally, the interactive data calculation submodule includes: The first calculation unit is used to input the biometric verification feature into the following formula to obtain a biometric verification score: , where C is the biometric verification score, n is the number of user feature types taken, s is the corresponding feature similarity, and w is the weight coefficient corresponding to the feature; The second calculation unit is used to input the question-answer verification reply data into the following formula to obtain the question-answer verification score: , where A refers to the question-answering verification score of selecting the template with the minimum edit distance ED given a user question Q and a set of predefined answer templates (T={t_1, t_2, ..., t_n}); The interactive data calculation unit is used to generate user interactive data based on the biometric verification score and the question-and-answer verification score.

[0057] Optional risk control assessment modules include: The weight acquisition submodule is used to obtain the active weight, passive weight and interactive weight corresponding to the user's active data, user's passive data and user's interactive data; The risk control calculation submodule is used to weight the user's active data, active weight, user's passive data, passive weight, user's interactive data and interactive weight to obtain the user's risk assessment score.

[0058] For the specific definition of the credit risk assessment device based on artificial intelligence, please refer to the definition of the credit risk assessment method based on artificial intelligence above, which will not be repeated here. Each module in the above-mentioned credit risk assessment device based on artificial intelligence can be implemented in whole or in part by software, hardware and their combination. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0059] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a credit risk assessment method based on artificial intelligence is implemented.

[0060] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: Obtain the credit application request triggered by the user, obtain the user's basic information from the credit application request, and extract the user's active data from the user's basic data; Obtain the user's current location from the credit application request, and perform calculations based on the user's current location to obtain the user's passive data; Generate and trigger a user interaction verification request to the user, and obtain corresponding user interaction data according to the user interaction verification request; Calculations are performed based on user active data, user passive data, and user interaction data to obtain a user risk assessment score, and user credit risk control results are generated based on the user risk assessment score.

[0061] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: Obtain the credit application request triggered by the user, obtain the user's basic information from the credit application request, and extract the user's active data from the user's basic data; Obtain the user's current location from the credit application request, and perform calculations based on the user's current location to obtain the user's passive data; Generate and trigger a user interaction verification request to the user, and obtain corresponding user interaction data according to the user interaction verification request; Calculations are performed based on user active data, user passive data, and user interaction data to obtain a user risk assessment score, and user credit risk control results are generated based on the user risk assessment score.

[0062] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0063] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0064] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A credit risk assessment method based on artificial intelligence, characterized in that: The credit risk assessment method based on artificial intelligence includes: Obtaining a credit application request triggered by a user, obtaining basic user information from the credit application request, and extracting active user data from the basic user data; Obtaining the user's current location from the credit application request, and performing calculations based on the user's current location to obtain user passive data; Generate and trigger a user interaction verification request to the user, and obtain corresponding user interaction data according to the user interaction verification request; Calculation is performed based on the user active data, user passive data and user interactive data to obtain a user risk assessment score, and a user credit risk control result is generated based on the user risk assessment score.

2. The credit risk assessment method based on artificial intelligence according to claim 1, characterized in that: The step of obtaining the user's current location from the credit application request and performing calculations based on the user's current location to obtain user passive data specifically includes: Obtaining preset high-risk areas, and obtaining the center position of each of the high-risk areas as the corresponding high-risk position; The risk distance between the user's current location and each of the high-risk locations is obtained, and the risk distance with the smallest value is input into the following formula to calculate the user's passive data: , where R s refers to the user passive data, d refers to the minimum risk distance, σ refers to the preset risk curve coefficient, and k refers to the preset risk coefficient.

3. The credit risk assessment method based on artificial intelligence according to claim 1, characterized in that: The generating and triggering a user interaction verification request to the user, and obtaining corresponding user interaction data according to the user interaction verification request specifically includes: The user interaction verification request includes a biometric verification request and a question-and-answer verification request, and the biometric verification feature and the question-and-answer verification reply data are obtained from the biometric verification request and the question-and-answer verification request respectively; The biometric verification feature and the question-and-answer verification response data are calculated respectively to obtain the user interaction data.

4. The credit risk assessment method based on artificial intelligence according to claim 3, characterized in that: The calculating the biometric verification feature and the question-and-answer verification response data respectively to obtain the user interaction data specifically includes: The biometric verification feature is input into the following formula to obtain the biometric verification score: , where C is the biometric verification score, n is the number of user feature types taken, s is the corresponding feature similarity, and w is the weight coefficient corresponding to the feature; The question-and-answer verification response data is input into the following formula to obtain the question-and-answer verification score: , where A refers to the question-answering verification score of selecting the template with the minimum edit distance ED given a user question Q and a set of predefined answer templates (T={t_1, t_2, ..., t_n}); The user interaction data is generated according to the biometric verification score and the question-and-answer verification score.

5. The credit risk assessment method based on artificial intelligence according to any one of claims 1 to 4, characterized in that: The calculation is performed based on the user active data, the user passive data and the user interactive data to obtain a user risk assessment score, and the user credit risk control result is generated based on the user risk assessment score, specifically including: Obtaining active weights, passive weights and interactive weights corresponding to the user active data, the user passive data and the user interactive data; The user active data, the active weight, the user passive data, the passive weight, the user interactive data and the interactive weight are weighted to obtain the user risk assessment score.

6. A credit risk assessment device based on artificial intelligence, characterized in that: The credit risk assessment device based on artificial intelligence includes: An active data acquisition module, used to acquire a credit application request triggered by a user, acquire user basic information from the credit application request, and extract user active data from the user basic data; A passive data acquisition module, used to acquire the user's current location from the credit application request, and perform calculations based on the user's current location to obtain user passive data; An interaction data acquisition module, used to generate and trigger a user interaction verification request to the user, and acquire corresponding user interaction data according to the user interaction verification request; The risk control assessment module is used to calculate based on the user's active data, user's passive data and user's interactive data to obtain a user risk assessment score, and generate a user credit risk control result based on the user's risk assessment score.

7. The artificial intelligence-based credit risk assessment device according to claim 6, characterized in that: The passive data acquisition module comprises: The risk location acquisition submodule is used to acquire preset high-risk areas and acquire the center position of each high-risk area as the corresponding high-risk location; The passive data calculation submodule is used to obtain the risk distance between the user's current location and each of the high-risk locations, and input the risk distance with the smallest value into the following formula to calculate the user's passive data: , where R s refers to the user passive data, d refers to the minimum risk distance, σ refers to the preset risk curve coefficient, and k refers to the preset risk coefficient.

8. The artificial intelligence-based credit risk assessment device according to claim 6, characterized in that: The interactive data acquisition module includes: An interactive data acquisition submodule, used for the user interactive verification request including a biometric verification request and a question-and-answer verification request, to respectively acquire the biometric verification feature and the question-and-answer verification reply data from the biometric verification request and the question-and-answer verification request; The interactive data calculation submodule is used to calculate the biometric verification feature and the question-and-answer verification response data respectively to obtain the user interactive data.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the credit risk assessment method based on artificial intelligence as claimed in any one of claims 1 to 5 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the credit risk assessment method based on artificial intelligence as claimed in any one of claims 1 to 5 are implemented.

Images