Artificial intelligence-based unhealthy asset assessment method and system

Abstract

The invention provides an artificial intelligence-based non-performing asset assessment method and system, and belongs to the technical field of data processing, and the method specifically comprises the steps: determining the deviation between the user data of a target user in a non-performing asset package and the user data of the target user in other non-performing asset packages, and determining the user data of the target user in the non-performing asset package; and determining a data deviation user in the target users by using the deviation condition and the updating condition of the user data, and determining the data deviation user in the target users when the data deviation user in the non-performing asset package meets the requirement by using an artificial intelligence model according to the non-performing asset data corresponding to the data deviation user in the non-performing asset package. According to the method, the evaluation analysis result of the unhealthy asset package is determined on the basis of deviation conditions of different target users, updating conditions of user data and the unhealthy asset data, differential evaluation of the unhealthy asset package is achieved, and meanwhile accurate evaluation of the data credibility of the unhealthy asset package is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and in particular relates to a non-performing asset evaluation method and system based on artificial intelligence. Background Art

[0002] Strictly speaking, non-performing assets are also called non-performing claims, the most important of which are non-performing loans, which refer to loans that the borrower cannot repay on time and in full. In other words, the loans issued by financial institutions cannot recover the principal and interest according to the pre-agreed term and interest rate. Therefore, how to manage non-performing assets and improve the asset recovery rate of non-performing assets has become a technical problem that needs to be solved urgently.

[0003] Specifically, in the existing technical solutions CN202311473599.6 "A method for valuing non-performing assets based on multimodal high-dimensional features", CN202111560706.X "A method and device for evaluating a non-performing asset package", and CN202211385059.8 "A method for evaluating overdue assets in consumer finance based on big data", similar technical means for evaluating and managing non-performing assets are given, thereby greatly improving the asset recovery rate of non-performing assets. However, the above technical solutions have the following problems: In the process of evaluating non-performing assets, the authenticity of the data of non-performing assets is ignored in the existing technical solutions. As time goes by, the user data corresponding to the non-performing assets will inevitably change, thereby affecting the authenticity of the user data corresponding to the non-performing assets. Therefore, if the authenticity of the user data corresponding to the non-performing assets is ignored, it is impossible to ensure that the income from the recovery and disposal of non-performing assets meets the requirements.

[0004] In response to the above technical problems, this application specifically provides an artificial intelligence-based non-performing asset assessment method and system. Summary of the invention

[0005] To achieve the purpose of the present invention, the present invention adopts the following technical solutions: In the first aspect, the present application provides a method for evaluating non-performing assets based on artificial intelligence, which specifically includes: S1: Perform parsing of the non-performing asset package, determine the update status of the user data of the target user in the non-performing asset package by using the parsing result of the parsing, and when it is determined that the real-time performance of the data of the non-performing asset package can meet the requirements in combination with the non-performing asset data of the target user, proceed to the next step; S2: Determine the deviation between the user data of the target user in the non-performing asset package and the user data of the target user in other non-performing asset packages, and determine the data deviation user among the target users by using the deviation and the update status of the user data; S3: According to the non-performing asset data corresponding to the data deviation users in the non-performing asset package, an artificial intelligence model is used to determine that when the data deviation users in the non-performing asset package meet the requirements, the evaluation and analysis results of the non-performing asset package are determined based on the deviation conditions of different target users, the update status of user data and the non-performing asset data.

[0006] The beneficial effects of the present invention are: 1. Utilize the update status of user data of target users in the non-performing asset package and the non-performing asset data of target users to determine whether the real-time performance of the data of the non-performing asset package can meet the requirements. This not only takes into account the difference in the real-time performance of user data of target users due to the difference in the update time of user data, but also takes into account the impact on the difficulty of debt collection and processing of the non-performing asset package when there is a deviation in the real-time performance of user data due to the difference in the non-performing asset data of target users. This realizes the screening of non-performing asset packages whose data real-time performance does not meet the requirements, and improves the processing efficiency of the evaluation and analysis of non-performing asset packages in some cases.

[0007] 2. The evaluation and analysis results of the non-performing asset package are determined based on the deviations of different target users, the update status of user data and the non-performing asset data, taking full account of the data update time of different target users and the data deviations from other non-performing asset packages, thereby achieving a comprehensive evaluation of the credibility of the target users' user data from multiple dimensions, avoiding the technical problem of excessive difficulty in debt collection due to a large number of target users with low data credibility, and ensuring the accuracy of the evaluation and analysis results of the non-performing asset package.

[0008] A further technical solution is that the update status of the user data of the target user includes the update time of the user data of different dimensions of the target user.

[0009] A further technical solution is that the non-performing asset data of the target user includes the overdue amount of the target user.

[0010] A further technical solution is to determine whether the real-time performance of the data of the non-performing asset package can meet the requirements, which specifically includes: Determine the update time of the user data of the target user in different dimensions based on the update status of the user data of the target user, and use the update time to determine the update delayed users among the target users; Based on the non-performing asset data of different update delay users, determine the sum of the overdue amounts of different update delay users, and use it as the update delay overdue amount; By measuring the proportion of the updated delayed overdue amount in the non-performing asset package, it is determined whether the real-time performance of the data of the non-performing asset package can meet the requirements.

[0011] A further technical solution is that the update delayed user is a target user whose update time is not within a preset time range and whose user data dimension number is greater than a preset dimension number.

[0012] A further technical solution is that the method for determining the evaluation and analysis results of the non-performing asset package is: Determine the update delay dimension and the update reliability dimension in the user data of different target users based on the update status of the user data of different target users, and determine the data deviation dimension in the update reliability dimension based on the deviation status in the update reliability dimension of the target users; Determine the data deviation weight values ​​of different target users by using the update delay dimension and data deviation dimension ratio of different target users; According to the non-performing asset data of different target users, the proportion of the overdue amount of different target users in the non-performing asset package is determined, and used as the overdue weight coefficient. Based on the sum of the data deviation weight values ​​of different target users and the product of the overdue weight coefficient, the asset package data deviation value of the non-performing asset package is determined, and the asset package data deviation value of the non-performing asset package is used to determine the evaluation and analysis results of the non-performing asset package.

[0013] A further technical solution is to determine the evaluation and analysis result of the non-performing asset package by using the asset package data deviation value of the non-performing asset package, which specifically includes: When the asset package data deviation value of the non-performing asset package is not within the preset deviation value range, the non-performing asset package is determined to be a Class I asset package; When the asset package data deviation value of the non-performing asset package is within a preset deviation value interval, the non-performing asset package is regarded as a second-class asset package.

[0014] A further technical solution is that the data validity of the first type of asset package is greater than that of the second type of asset package.

[0015] In a second aspect, the present invention provides an artificial intelligence-based non-performing asset evaluation system, which adopts the above-mentioned artificial intelligence-based non-performing asset evaluation method, specifically comprising: Real-time evaluation module, deviation user identification module, analysis result output module; The real-time evaluation module is responsible for determining whether the real-time performance of the data of the non-performing asset package can meet the requirements; The deviation user identification module is responsible for determining data deviation users among target users; The analysis result output module is responsible for determining the evaluation and analysis results of the non-performing asset package.

[0016] Other features and advantages will be described in the following description. The objects and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description and drawings.

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.

[0019] Figure 1 It is a flowchart of a method for evaluating non-performing assets based on artificial intelligence; Figure 2 It is a flow chart to determine whether the real-time performance of the data of the non-performing asset package can meet the requirements; Figure 3 is a flow chart of a method for determining a data deviation user among target users; Figure 4 It is a flow chart of the method for determining the evaluation and analysis results of the non-performing asset package; Figure 5 It is a framework diagram of an artificial intelligence-based non-performing asset evaluation system. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0021] In this application, the credibility of the user data in the non-performing asset package is evaluated to obtain the evaluation and analysis results of the non-performing asset package, which fully considers the difference in the difficulty of subsequent collection processing due to the deviation of the user data in the non-performing asset package.

[0022] Target users whose update time is not within the preset time range and whose user data dimension number is greater than the preset dimension number are taken as update delay users, and the sum of the overdue amounts of different update delay users is determined and taken as the update delay overdue amount. When the proportion of the update delay overdue amount in the overdue amount in the non-performing asset package is greater than 0.3, it is determined that the real-time performance of the data of the non-performing asset package cannot meet the requirements.

[0023] Determine the update delay dimension and update reliability dimension in the user data of the target user based on whether the update time of the user data of different target users is within the preset time range, and use the deviation between the update reliability dimension of the target user and the user data of other non-performing asset packages as the data deviation dimension; When the proportion of the update delay dimension and the data deviation dimension of the target user is greater than 0.6, the target user is determined to be a data deviation user.

[0024] The non-performing asset data of the data deviation user and the non-performing asset data of the non-performing asset package are used as the input of the artificial intelligence model, and according to the output of the artificial intelligence model, it is determined whether the data deviation user in the non-performing asset package meets the requirements.

[0025] The data deviation weight values ​​of different target users are determined by utilizing the quantitative proportions of the update delay dimensions and data deviation dimensions of different target users; the proportions of the overdue amounts of different target users in the non-performing asset package are determined according to the non-performing asset data of different target users, and used as the overdue weight coefficient; the asset package data deviation value of the non-performing asset package is determined based on the sum of the products of the data deviation weight values ​​of different target users and the overdue weight coefficient; when the asset package data deviation value of the non-performing asset package is not within the preset deviation value range, the non-performing asset package is determined to be a Class I asset package; when the asset package data deviation value of the non-performing asset package is within the preset deviation value range, the non-performing asset package is determined to be a Class II asset package.

[0026] Example 1 like Figure 1 As shown, in the first aspect, the present application provides a non-performing asset evaluation method based on artificial intelligence, which specifically includes: S1: Perform parsing of the non-performing asset package, determine the update status of the user data of the target user in the non-performing asset package by using the parsing result of the parsing, and when it is determined that the real-time performance of the data of the non-performing asset package can meet the requirements in combination with the non-performing asset data of the target user, proceed to the next step; Furthermore, the update status of the user data of the target user includes the update time of the user data of different dimensions of the target user.

[0027] Specifically, the non-performing asset data of the target user includes the overdue amount of the target user.

[0028] It should be noted that if Figure 2 As shown, it is determined that the real-time performance of the data of the non-performing asset package can meet the requirements, specifically including: S11: determining the update time of the user data of the target user in different dimensions based on the update status of the user data of the target user, and determining the update-delayed users among the target users by using the update time; S12: Based on the non-performing asset data of different update delay users, determine the sum of the overdue amounts of different update delay users, and use it as the update delay overdue amount; S13: Determine whether the real-time performance of the data of the non-performing asset package can meet the requirements by calculating the proportion of the updated delayed overdue amount in the non-performing asset package.

[0029] Furthermore, the update delayed user is a target user whose update time is not within a preset time range and whose user data dimension number is greater than a preset dimension number.

[0030] Optionally, when the proportion of the update delayed overdue amount in the non-performing asset package is greater than the preset amount proportion, it is determined that the real-time performance of the data of the non-performing asset package cannot meet the requirements.

[0031] Furthermore, when the real-time performance of the data of the non-performing asset package cannot meet the requirements, the evaluation and analysis result of the non-performing asset package is determined to be a Class I asset package.

[0032] In another possible embodiment, determining whether the real-time performance of the data of the non-performing asset package can meet the requirements specifically includes: Determine the update time of the user data of the target user in different dimensions based on the update status of the user data of the target user, use the update time to determine the proportion of the number of dimensions of the user data of the target user whose update time is not within the preset time range, and use the proportion as the update delay coefficient of the target user; Based on the non-performing asset data of different target users, determine the proportion of the overdue amount of different target users in the non-performing asset package, and use it as the proportion of the overdue amount of the target user; The product of the update delay coefficient and the overdue amount ratio of the different target users is summed to determine the user data delay of the non-performing asset package, and the user data delay is used to determine whether the real-time performance of the data of the non-performing asset package can meet the requirements.

[0033] Furthermore, the user data delay of the non-performing asset package has a value range of 0 to 1, wherein when the user data delay of the non-performing asset package is less than a preset delay threshold, it is determined that the data real-time performance of the non-performing asset package can meet the requirements.

[0034] In another possible embodiment, determining whether the real-time performance of the data of the non-performing asset package can meet the requirements specifically includes: Based on the update status of the user data of the target user, the update time of the user data of different dimensions of the target user is determined. When the update time of the user data of different target users is determined to be within a preset time range, it is determined that the real-time performance of the data of the non-performing asset package can meet the requirements; When the update time is used to determine that there is a target user whose user data update time is not within the preset time: The update time is used to determine the update time of user data of different target users in the specified dimension. When the update time of user data in the specified dimension is not within the preset time range and the proportion of the number of target users does not meet the requirement, it is determined that the real-time performance of the data of the non-performing asset package can meet the requirement; When the percentage of target users whose user data in a specified dimension is updated outside the preset time range meets the requirement: Determine the update delay users among the target users by using the update time, determine the sum of the overdue amounts of different update delay users based on the non-performing asset data of different update delay users, and use it as the update delay overdue amount, and when the proportion of the update delay overdue amount in the overdue amount of the non-performing asset package does not meet the requirement, determine that the data real-time performance of the non-performing asset package cannot meet the requirement; When the proportion of the overdue amount of the update delay in the good asset package meets the requirements: The proportion of the number of dimensions of the user data of the target user whose update time is not within the preset time range is used as the update delay coefficient of the target user; Based on the non-performing asset data of different target users, determine the proportion of the overdue amount of different target users in the non-performing asset package, and use it as the proportion of the overdue amount of the target user; The product of the update delay coefficient and the overdue amount ratio of the different target users is summed to determine the user data delay of the non-performing asset package, and the user data delay is used to determine whether the real-time performance of the data of the non-performing asset package can meet the requirements.

[0035] Furthermore, the user data of the specified dimension includes the user's home address, the user's contact information, and the user's company address.

[0036] S2: Determine the deviation between the user data of the target user in the non-performing asset package and the user data of the target user in other non-performing asset packages, and determine the data deviation user among the target users by using the deviation and the update status of the user data; It can be understood that the deviation situation includes other non-performing asset packages that have deviations from the user data of the target user in the non-performing asset package and other non-performing asset packages that are consistent with the user data of the target user in the non-performing asset package.

[0037] Specifically, Figure 3 As shown, the method for determining the data deviation users among the target users is: S201: determining the update time of the user data of the target user in different dimensions according to the user data of the target user in different dimensions, and determining the proportion of the dimensions of the user data whose update time is not within a preset time range, and using it as an update delay coefficient; S202: Based on the deviation, determine the deviation of user data in different dimensions from user data of other non-performing asset packages, and determine the data deviation dimension in the dimension of the user data by using the number of other non-performing asset packages with deviations, and determine the data deviation coefficient by using the proportion of the number of dimensions of the data deviation dimension; S203: Taking an average value of the data deviation coefficient and the updated delay coefficient as the delay deviation coefficient of the target user, and using the delay deviation coefficient to determine whether the target user is a data deviation user.

[0038] Further, when the delay deviation coefficient of the target user is not within a preset coefficient range, the target user is determined to be a data deviation user.

[0039] It can be understood that the method for determining the data deviation users among the target users is: Determine the update time of the user data in different dimensions according to the user data of the target user in different dimensions, and determine the dimension of the user data whose update time is not within the preset time range, and use it as the update delay dimension; Based on the deviation, determine the deviation of user data in different dimensions from user data of other non-performing asset packages, and determine the data deviation dimension in the dimension of the user data by using the number of other non-performing asset packages with deviations; Whether the target user is a data deviation user is determined according to the sum of the update delay dimension and the data deviation dimension of the target user.

[0040] Further, when the sum of the update delay dimension and the data deviation dimension of the target user does not meet the requirement, the target user is determined to be a data deviation user.

[0041] In another possible embodiment, a method for determining data deviation users among the target users is: S21: determining the update time of the user data of the target user in different dimensions according to the user data of the target user in different dimensions, and determining the dimension of the user data whose update time is not within a preset time range, taking it as the update delay dimension, and determining the update delay coefficient of the target user according to the proportion of the number of dimensions of the update delay dimension; S22: Based on the deviation, determine the deviation between the user data of different dimensions and the user data of other non-performing asset packages, and determine the user data deviation value of the target user by using the number of other non-performing asset packages with deviations in the user data of different dimensions and the number of consistent non-performing asset packages; S23: Determine a data delay deviation value of the target user according to the update delay coefficient and the user data deviation value of the target user, and use the data delay deviation value to determine whether the target user is a data deviation user.

[0042] Optionally, the above step S21 includes the following contents: S211: determining the update time of the user data of the target user in different dimensions according to the user data of the target user in different dimensions, and determining the dimension of the user data whose update time is not within a preset time range, and using it as the update delay dimension; S212: Determine the update delay coefficient of the target user using the proportion of the number of dimensions of the update delay dimension; when the update delay coefficient of the target user is greater than a preset update delay coefficient threshold, determine that the target user is a data deviation user; when the update delay coefficient of the target user is not greater than the preset update delay coefficient threshold, proceed to step S22.

[0043] Optionally, the above step S22 includes the following contents: S221: Based on the deviation, determine the deviation between the user data of different dimensions and the user data of other non-performing asset packages. When the user data of different dimensions do not deviate from the user data of other non-performing asset packages, determine that the target user does not belong to the data deviation user. When there is user data that deviates from the user data of other non-performing asset packages, proceed to step S222. S222: Determine the data deviation dimension in the dimension of the user data by using the number of other non-performing asset packages with deviations. When the number ratio of the data deviation dimension of the target user does not meet the requirement, determine that the target user is a data deviation user. When the number ratio of the data deviation dimension of the target user meets the requirement, proceed to step S223. S223: using the user data excluding the update delay dimension as the update reliability dimension, and when the number ratio of the data deviation dimension in the update reliability dimension does not meet the requirement, determining that the target user is a data deviation user, and when the number ratio of the data deviation dimension in the update reliability dimension meets the requirement, proceeding to step S224; S224: Obtain the proportion of the number of data deviation dimensions in the update reliable dimension, and determine the data deviation amount of the update reliable dimension in combination with the number of deviations between different update reliable dimensions and other non-performing asset packages. When the data deviation amount of the update reliable dimension does not meet the requirement, it is determined that the target user belongs to a data deviation user. When the data deviation amount of the update reliable dimension meets the requirement, proceed to step S225; S225: Determine the user data deviation value of the target user by using the number of other non-performing asset packages with deviations in user data of different dimensions and the number of consistent non-performing asset packages. When the user data deviation value of the target user does not meet the requirements, it is determined that the target user is a data deviation user. When the user data deviation value of the target user meets the requirements, proceed to step S23.

[0044] S3: According to the non-performing asset data corresponding to the data deviation users in the non-performing asset package, an artificial intelligence model is used to determine that when the data deviation users in the non-performing asset package meet the requirements, the evaluation and analysis results of the non-performing asset package are determined based on the deviation conditions of different target users, the update status of user data and the non-performing asset data.

[0045] Further, determining whether the data deviation in the non-performing asset package meets the requirements includes: The non-performing asset data of the data deviation user and the non-performing asset data of the non-performing asset package are used as the input of the artificial intelligence model, and according to the output of the artificial intelligence model, it is determined whether the data deviation user in the non-performing asset package meets the requirements.

[0046] It should be noted that the artificial intelligence model is constructed using any one or more of a BP neural network, a LSTM neural network, a RNN neural network and a general large model.

[0047] Specifically, when the output of the artificial intelligence model is 1, it is determined that the data deviation in the non-performing asset package does not meet the user requirements; when the output of the artificial intelligence model is 0, it is determined that the data deviation in the non-performing asset package meets the user requirements.

[0048] It is understandable that when the data deviation in the non-performing asset package does not meet the user's requirements, the evaluation and analysis result of the non-performing asset package is determined to be a Class I asset package.

[0049] Specifically, Figure 4 As shown, the method for determining the evaluation and analysis results of the non-performing asset package is: S31: determining an update delay dimension and an update reliability dimension in the user data of different target users based on the update status of the user data of different target users, and determining a data deviation dimension in the update reliability dimension based on the deviation status in the update reliability dimension of the target users; S32: Determine data deviation weight values ​​for different target users by using the proportion of update delay dimensions and data deviation dimensions for different target users; S33: Determine the proportion of the overdue amount of different target users in the non-performing asset package based on the non-performing asset data of different target users, and use it as the overdue weight coefficient; determine the asset package data deviation value of the non-performing asset package based on the sum of the product of the data deviation weight value of different target users and the overdue weight coefficient; and use the asset package data deviation value of the non-performing asset package to determine the evaluation and analysis result of the non-performing asset package.

[0050] Further, the evaluation and analysis results of the non-performing asset package are determined by using the asset package data deviation value of the non-performing asset package, specifically including: When the asset package data deviation value of the non-performing asset package is not within the preset deviation value range, the non-performing asset package is determined to be a Class I asset package; When the asset package data deviation value of the non-performing asset package is within a preset deviation value interval, the non-performing asset package is regarded as a second-class asset package.

[0051] It should be noted that the data validity of the first category asset package is greater than that of the second category asset package.

[0052] Optionally, the method for determining the evaluation and analysis results of the non-performing asset package is: Determine the update delay dimension and the update reliability dimension in the user data of different target users based on the update status of the user data of different target users, and determine the data deviation dimension in the update reliability dimension based on the deviation status in the update reliability dimension of the target users; When the proportion of the number of target users corresponding to the update delay dimension and the data deviation dimension in the specified dimension does not meet the requirements, it is determined that the evaluation and analysis result of the non-performing asset package is a type I asset package; When the percentage of target users corresponding to the update delay dimension and data deviation dimension in different specified dimensions meets the requirements: Obtaining an average value of the number of target users corresponding to the update delay dimension and the data deviation dimension in different specified dimensions, when the average value of the number of target users corresponding to the update delay dimension and the data deviation dimension in different specified dimensions is greater than a preset number ratio threshold, determining that the evaluation and analysis result of the non-performing asset package is a Class I asset package; When the average value of the target user quantity ratio corresponding to the update delay dimension and the data deviation dimension in different specified dimensions is not greater than the preset quantity ratio threshold: The data deviation weight values ​​of different target users are determined by using the update delay dimension and the quantity proportion of the data deviation dimension of different target users. When the average value of the data deviation weight values ​​of different target users does not meet the requirements, the evaluation and analysis result of the non-performing asset package is determined to be a type I asset package. When the average value of the data deviation weights of different target users meets the requirements: Determine the proportion of the overdue amount of different target users in the non-performing asset package according to the non-performing asset data of different target users, and use it as the overdue weight coefficient, and divide the target users into different deviation weight value intervals using the data deviation weight value, and when the sum of the overdue weight coefficients within the preset deviation weight value interval does not meet the requirements, determine that the evaluation and analysis result of the non-performing asset package is a Class I asset package; When the sum of the overdue weight coefficients within the preset deviation weight value range meets the requirements: Based on the sum of the product of the data deviation weight values ​​of different target users and the overdue weight coefficient, the asset package data deviation value of the non-performing asset package is determined, and the asset package data deviation value of the non-performing asset package is used to determine the evaluation and analysis result of the non-performing asset package.

[0053] Furthermore, the preset deviation weight value interval is an interval in which the deviation weight value is greater than the preset deviation weight value.

[0054] Example 2 Second, as Figure 5 As shown, the present invention provides a non-performing asset evaluation system based on artificial intelligence, which adopts the above-mentioned non-performing asset evaluation method based on artificial intelligence, and specifically includes: Real-time evaluation module, deviation user identification module, analysis result output module; The real-time evaluation module is responsible for determining whether the real-time performance of the data of the non-performing asset package can meet the requirements; The deviation user identification module is responsible for determining data deviation users among target users; The analysis result output module is responsible for determining the evaluation and analysis results of the non-performing asset package.

[0055] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0056] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0057] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. A method for evaluating non-performing assets based on artificial intelligence, characterized in that: Specifically include: Performing parsing of the non-performing asset package, using the parsing result of the parsing to determine the update status of the user data of the target user in the non-performing asset package, and combining the non-performing asset data of the target user to determine that the real-time performance of the data of the non-performing asset package can meet the requirements, proceeding to the next step; Determine the deviation between the user data of the target user in the non-performing asset package and the user data of the target user in other non-performing asset packages, and determine the data deviation user among the target users by using the deviation and the update of the user data; According to the non-performing asset data corresponding to the data deviation users in the non-performing asset package, an artificial intelligence model is used to determine whether the data deviation users in the non-performing asset package meet the requirements. Then, the evaluation and analysis results of the non-performing asset package are determined based on the deviation conditions of different target users, the update status of user data and the non-performing asset data.

2. The artificial intelligence-based non-performing asset evaluation method according to claim 1, characterized in that: The update status of the user data of the target user includes the update time of the user data of different dimensions of the target user.

3. The artificial intelligence-based non-performing asset evaluation method according to claim 1, characterized in that: The non-performing asset data of the target user includes the overdue amount of the target user.

4. The artificial intelligence-based non-performing asset evaluation method according to claim 1, characterized in that: Determine whether the real-time performance of the data of the non-performing asset package can meet the requirements, including: Determine the update time of the user data of the target user in different dimensions based on the update status of the user data of the target user, and use the update time to determine the update delayed users among the target users; Based on the non-performing asset data of different update delay users, determine the sum of the overdue amounts of different update delay users, and use it as the update delay overdue amount; By measuring the proportion of the updated delayed overdue amount in the non-performing asset package, it is determined whether the real-time performance of the data of the non-performing asset package can meet the requirements.

5. The artificial intelligence-based non-performing asset evaluation method according to claim 4, characterized in that: The update delayed user is a target user whose update time is not within the preset time range and whose user data dimension number is greater than the preset dimension number.

6. The artificial intelligence-based non-performing asset evaluation method according to claim 4, characterized in that: When the real-time performance of the data of the non-performing asset package cannot meet the requirements, the evaluation and analysis result of the non-performing asset package is determined to be a Class I asset package.

7. The artificial intelligence-based non-performing asset evaluation method according to claim 1, characterized in that: When the data deviation in the non-performing asset package does not meet the user's requirements, the evaluation and analysis result of the non-performing asset package is determined to be a Class I asset package.

8. The artificial intelligence-based non-performing asset evaluation method according to claim 1, characterized in that: The method for determining the evaluation and analysis results of the non-performing asset package is: Determine the update delay dimension and the update reliability dimension in the user data of different target users based on the update status of the user data of different target users, and determine the data deviation dimension in the update reliability dimension based on the deviation status in the update reliability dimension of the target users; Determine the data deviation weight values ​​of different target users by using the update delay dimension and data deviation dimension ratio of different target users; According to the non-performing asset data of different target users, the proportion of the overdue amount of different target users in the non-performing asset package is determined, and used as the overdue weight coefficient. Based on the sum of the data deviation weight values ​​of different target users and the product of the overdue weight coefficient, the asset package data deviation value of the non-performing asset package is determined, and the asset package data deviation value of the non-performing asset package is used to determine the evaluation and analysis results of the non-performing asset package.

9. The artificial intelligence-based non-performing asset evaluation method according to claim 8, characterized in that: Determining the evaluation and analysis results of the non-performing asset package by using the asset package data deviation value of the non-performing asset package specifically includes: When the asset package data deviation value of the non-performing asset package is not within the preset deviation value range, the non-performing asset package is determined to be a Class I asset package; When the asset package data deviation value of the non-performing asset package is within a preset deviation value interval, the non-performing asset package is regarded as a second-class asset package.

10. An artificial intelligence-based non-performing asset evaluation system, using an artificial intelligence-based non-performing asset evaluation method according to any one of claims 1 to 9, characterized in that: Specifically include: Real-time evaluation module, deviation user identification module, analysis result output module; The real-time evaluation module is responsible for determining whether the real-time performance of the data of the non-performing asset package can meet the requirements; The deviation user identification module is responsible for determining data deviation users among target users; The analysis result output module is responsible for determining the evaluation and analysis results of the non-performing asset package.

Images