Credit card overdue processing method and device, equipment, medium and program product

Abstract

The invention provides a credit card overdue processing method and device, equipment, a medium and a program product, relates to the technical field of artificial intelligence, and can be applied to the field of financial science and technology. The method comprises the steps of obtaining multi-dimensional data of a target user under the condition that user authorization is obtained; constructing multi-dimensional features of the target user based on the multi-dimensional data of the target user; the multi-dimensional features of the target user are input into a trained credit card overdue user classification model to obtain a classification result output by the credit card overdue user classification model, and the classification result represents that the target user belongs to a first type of users or a second type of users; when the target user belongs to the first type of users, processing the credit card overdue behavior of the target user according to a first strategy; when the target user belongs to the second type of users, the credit card overdue behavior of the target user is processed according to a second strategy, and the second strategy is different from the first strategy.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology and can be used in the field of financial technology. More specifically, it relates to a method, apparatus, device, medium, and program product for processing credit card delinquency. Background Technology

[0002] Currently, for credit card delinquencies, financial institutions typically determine repayment strategies (such as installment periods and interest rates) based on the number of overdue days and the overdue amount, following fixed rules. However, this approach has significant limitations. First, it relies heavily on single-dimensional overdue indicators, such as the number of overdue days and the overdue amount, lacking in-depth analysis of user behavior patterns. For example, some users may experience temporary repayment difficulties due to unforeseen illnesses, unemployment, or other objective reasons, but the system may fail to identify these temporary delinquencies, leading to misjudgments, negatively impacting user experience, and increasing user churn. Second, using fixed installment templates when formulating repayment strategies fails to adequately consider the user's dynamic financial situation, such as income fluctuations and changes in debt. This lack of adaptability can result in repayment plans that do not match the user's actual repayment ability, thereby reducing the success rate of credit card delinquency repayments and increasing the risk of bad debts. Summary of the Invention

[0003] This invention provides a method, apparatus, device, medium, and program product for handling credit card delinquency, which can at least partially solve one or more of the above-mentioned problems.

[0004] A first aspect of this invention provides a method for handling credit card delinquency. The method includes: when the credit card delinquency data of a target user meets credit card delinquency detection conditions, obtaining user authorization to use multi-dimensional data of the target user; upon obtaining the user authorization, obtaining the multi-dimensional data of the target user; constructing multi-dimensional features of the target user based on the multi-dimensional data, the multi-dimensional features including behavioral features, financial features, and social relationship features; inputting the multi-dimensional features of the target user into a trained credit card delinquency user classification model to obtain a classification result output by the credit card delinquency user classification model, the classification result indicating whether the target user belongs to a first type of user or a second type of user; and when the target user belongs to the first type of user, processing the target user's credit card delinquency behavior according to a first strategy; and when the target user belongs to the second type of user, processing the target user's credit card delinquency behavior according to a second strategy, wherein the second strategy is different from the first strategy.

[0005] According to an embodiment of the present invention, when the target user belongs to the second type of user, the step of handling the target user's credit card delinquency behavior according to the second strategy includes: generating a personalized repayment plan for the target user based on at least some data of the target user, macroeconomic data, and employment data of the target user's industry by using a strategy optimization model constructed based on reinforcement learning. The personalized repayment plan includes installment periods, tiered interest rates, and a temporary extension mechanism, wherein at least some data of the target user comes from the target user's multi-dimensional data.

[0006] According to an embodiment of the present invention, the strategy optimization model is a reinforcement learning model constructed using a Markov decision process. Generating a personalized repayment plan for the target user includes: the strategy optimization model simulating different repayment strategies within the Markov decision process framework to obtain the cumulative rewards of different repayment strategies; wherein at least one of the installment period, tiered interest rate, and temporary extension mechanism differs among the different repayment strategies; based on the cumulative rewards of the different repayment strategies, the execution effect of the different repayment strategies is obtained; and the repayment strategy with the best execution effect is selected from the different repayment strategies as the personalized repayment plan for the target user.

[0007] According to an embodiment of the present invention, obtaining the execution effect of different repayment strategies based on the cumulative rewards of different repayment strategies includes: verifying the execution success rate of different repayment strategies through Monte Carlo simulation; and determining the execution effect of each repayment strategy based on the cumulative rewards and the execution success rate of each repayment strategy.

[0008] According to an embodiment of the present invention, when the target user belongs to the second type of user, the step of handling the target user's credit card delinquency behavior according to the second strategy further includes: updating the multi-dimensional data of the target user in response to a preset period or preset event; constructing the current time series data of the target user in response to the update of the multi-dimensional data of the target user; inputting the time series data of the target user into a trained real-time risk warning model to obtain the user type conversion probability output by the real-time risk warning model, wherein the real-time risk warning model is a time series prediction model based on a long short-term memory network; and triggering an early warning that the target user is converted into a first type of user when the user type conversion probability is higher than a preset probability threshold.

[0009] According to an embodiment of the present invention, the method further includes: when the user type conversion probability is higher than a preset probability threshold, extracting multiple quantitative indicators based on the multi-dimensional data of the target user; when any one of the quantitative indicators exceeds the indicator threshold corresponding to that indicator, determining that the target user is converted into a first type of user; wherein the indicator thresholds corresponding to different quantitative indicators are different.

[0010] According to an embodiment of the present invention, obtaining the multi-dimensional data of the target user includes: obtaining the multi-dimensional data of the target user from a blockchain network. The method further includes: storing the classification result and the strategy corresponding to the classification result in the blockchain network, wherein the strategy corresponding to the classification result is either the first strategy or the second strategy.

[0011] A second aspect of this invention provides a device for processing overdue credit card payments. The device includes: a data acquisition module, a feature engineering module, a user classification module, and a processing module.

[0012] The data acquisition module is used to obtain user authorization to use the multi-dimensional data of the target user when the target user's credit card delinquency data meets the credit card delinquency detection conditions; and to obtain the multi-dimensional data of the target user when the user authorization is obtained.

[0013] The feature engineering module is used to construct multidimensional features of the target user based on the target user's multidimensional data. The multidimensional features include behavioral features, financial features, and social relationship features.

[0014] The user classification module is used to input the multidimensional features of the target user into the trained credit card delinquency user classification model to obtain the classification result output by the credit card delinquency user classification model. The classification result indicates that the target user belongs to the second type of user or the first type of user.

[0015] The processing module is used to process the credit card delinquency behavior of the target user according to a first strategy when the target user belongs to a first type of user; and to process the credit card delinquency behavior of the target user according to a second strategy when the target user belongs to a second type of user, wherein the second strategy is different from the first strategy.

[0016] According to an embodiment of the present invention, the apparatus further includes a dynamic repayment strategy generation module. The dynamic repayment strategy generation module is used to generate a personalized repayment plan for the target user based on at least some data of the target user, macroeconomic data, and employment data of the target user's industry, using a strategy optimization model constructed based on reinforcement learning. The personalized repayment plan includes installment periods, tiered interest rates, and a temporary extension mechanism, wherein at least some data of the target user comes from the target user's multi-dimensional data.

[0017] According to an embodiment of the present invention, the device further includes a real-time early warning module. The real-time early warning module is configured to: update the multi-dimensional data of the target user in response to a preset period or a preset event; construct the current time-series data of the target user in response to the update of the multi-dimensional data of the target user; input the time-series data of the target user into a trained real-time risk early warning model to obtain the user type conversion probability output by the real-time risk early warning model, wherein the real-time risk early warning model is a time-series prediction model constructed based on a long short-term memory network; and trigger an early warning that the target user has been converted into a first-type user when the user type conversion probability is higher than a preset probability threshold.

[0018] A third aspect of the present invention provides an electronic device. The electronic device includes: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0019] A fourth aspect of the present invention also provides a computer-readable storage medium having a computer program or instructions stored thereon, wherein the computer program or instructions, when executed by a processor, implement the steps of the above-described method.

[0020] A fifth aspect of the present invention also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the steps of the above-described method. Attached Figure Description

[0021] The above-described features, other objects, and advantages of the present invention will become clearer from the following description of embodiments of the invention with reference to the accompanying drawings, in which:

[0022] Figure 1 The illustration schematically depicts an application scenario of a method, apparatus, device, medium, and program product for handling credit card delinquency according to embodiments of the present invention.

[0023] Figure 2 A flowchart illustrating a method for handling credit card delinquency according to an embodiment of the present invention is shown schematically;

[0024] Figure 3 This illustration shows a flowchart of generating personalized repayment plans for a second type of user in one embodiment of the present invention.

[0025] Figure 4 This illustration schematically shows a flowchart of monitoring the risk of a second type of user converting to a first type of user in another embodiment of the present invention;

[0026] Figure 5A flowchart illustrating a method for handling credit card delinquency according to another embodiment of the present invention is shown schematically;

[0027] Figure 6 A schematic diagram illustrating a structural block diagram of a credit card delinquency processing device according to an embodiment of the present invention is shown; and

[0028] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing a credit card delinquency processing method according to an embodiment of the present invention. Detailed Implementation

[0029] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the invention. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the invention for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0030] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0031] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0032] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0033] In the technical solution of this invention, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0034] In scenarios involving automated decision-making using personal information, the methods, devices, and systems provided in this invention offer users corresponding entry points for choosing to agree to or reject the automated decision-making results. If the user chooses to reject, the process proceeds to the expert decision-making stage. Here, "automated decision-making" refers to the activity of automatically analyzing and evaluating an individual's behavioral habits, interests, or economic, health, and credit status through computer programs, and then making a decision. Here, "expert decision-making" refers to the activity of making decisions by personnel who specialize in a particular field, possess specialized experience, knowledge, and skills, and have reached a certain level of professional expertise.

[0035] Embodiments of the present invention provide a method, apparatus, device, medium, and program product for handling credit card delinquency. According to embodiments of the present invention, a 360° user profile can be constructed by integrating multi-dimensional user characteristics (such as behavioral characteristics, financial characteristics, and social relationship characteristics), and a credit card delinquency user classification model can be built based on these multi-dimensional characteristics to classify users with credit card delinquency into a first category and a second category. The first category of users mainly refers to users who have the ability to repay but have delinquent credit card payments due to low willingness to repay. The second category of users mainly refers to users who are experiencing temporary repayment difficulties due to objective reasons.

[0036] Furthermore, this embodiment of the invention employs different strategies for subsequent processing based on the different reasons for delinquency between the first type of users and the second type of users. This not only improves the success rate of credit card delinquency repayment but also avoids adverse effects on users with temporary repayment difficulties, enhancing user experience and improving the credit card service capabilities of financial institutions.

[0037] Figure 1 The illustration schematically depicts an application scenario of a method, apparatus, device, medium, and program product for handling credit card delinquency according to embodiments of the present invention.

[0038] like Figure 1As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0039] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, instant messaging tools, email clients, social media platform software, e-banking, etc. (for example only).

[0040] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0041] Server 105 can be a server providing a banking system, capable of providing credit card transaction services. For example, it could be a backend service provided to an e-banking client accessed by a user via first terminal device 101, second terminal device 102, and third terminal device 103 (this is just an example).

[0042] It should be noted that the credit card overdue processing method provided in this embodiment of the invention can generally be executed by server 105. Correspondingly, the credit card overdue processing device provided in this embodiment of the invention can generally be located in server 105. The credit card overdue processing method provided in this embodiment of the invention can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the credit card overdue processing device provided in this embodiment of the invention can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0043] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0044] The following will be based on Figure 1 The described scene, through Figures 2-5 A method for handling credit card delinquency according to an embodiment of the present invention will be described in detail.

[0045] Figure 2 A flowchart illustrating a method for handling credit card delinquency according to an embodiment of the present invention is shown.

[0046] like Figure 2 As shown, the method 200 includes operations S210 to S240 and operations S251 or S252.

[0047] In operation S210, when the target user's credit card delinquency data meets the credit card delinquency detection conditions, user authorization to use the target user's multi-dimensional data is obtained. This multi-dimensional data may include the user's financial, work, health, and social data.

[0048] In operation S220, if the user authorization is obtained, multi-dimensional data of the target user is acquired.

[0049] In embodiments of the present invention, user consent or authorization can be obtained before acquiring multi-dimensional user data. For example, a request to acquire user information can be sent to the user before operation S220. If the user consents or authorizes the acquisition of user information, operation S220 is performed. Alternatively, the user's authorization to use their related data can be obtained when the user signs up for a credit card.

[0050] In operation S230, based on the multi-dimensional data of the target user, a multi-dimensional feature of the target user is constructed, which includes behavioral features, financial features, and social relationship features.

[0051] In operation S240, the multidimensional features of the target user are input into the trained credit card delinquency user classification model to obtain the classification result output by the credit card delinquency user classification model. The classification result indicates that the target user belongs to the first type of user or the second type of user.

[0052] Behavioral characteristics can include, for example, a user's spending patterns, repayment habits, and transaction frequency. Financial characteristics can cover income flow, debt ratio, and asset status. Social relationship characteristics can involve the stability of a user's social network and occupational changes. By integrating this multi-dimensional data to build a comprehensive user profile, it is possible to more accurately identify the first and second types of users, avoiding misjudgments caused by relying on a single indicator (such as the number of overdue days or the amount overdue).

[0053] In operation S251, when the target user belongs to the first category of users, the target user's credit card delinquency is handled according to the first strategy. The first strategy could be, for example, designating the target user as a blacklisted user and controlling their access to bank services, such as restricting high-end spending, reducing credit limits, or setting shorter repayment periods and higher interest rates.

[0054] In operation S252, when the target user belongs to the second type of user, the target user's credit card delinquency behavior is handled according to the second strategy, wherein the second strategy is different from the first strategy.

[0055] For example, for the second type of user, a longer grace period or a lower repayment interest rate can be set.

[0056] For example, for the second type of user, a personalized repayment plan can be generated based on the target user's actual situation (such as sudden unemployment, illness, etc.). The personalized repayment plan may include, but is not limited to: dynamically adjusted installment repayment periods (such as adaptive adjustment from 3 to 24 periods), tiered interest rates (such as the combination of base interest rate and risk premium), and temporary extension mechanisms (under what circumstances an extension can be applied for and for how long).

[0057] In one embodiment, a policy optimization model based on reinforcement learning is used to generate a personalized repayment plan for the target user, based on at least some of the target user's data, macroeconomic data, and employment data of the target user's industry. This improves the applicability of the repayment plan, increases the repayment success rate, and enhances the user experience. The personalized repayment plan includes installment periods, tiered interest rates, and temporary extension mechanisms. At least some of the target user's data comes from multi-dimensional data, such as real-time financial, work, and health data, or the health status of family members.

[0058] As can be seen, the embodiments of the present invention can classify credit card overdue users into first-class users and second-class users, and then adopt different strategies for different categories of users. This can not only improve the success rate of credit card overdue repayment, but also avoid adverse effects on users with temporary repayment difficulties, improve user experience, and enhance the credit card service capabilities of financial institutions.

[0059] Existing technologies typically rely on single dimensions such as overdue days and loan amounts for user identification and repayment plan determination. In contrast, this invention introduces multi-dimensional data, including behavioral characteristics, financial characteristics, and social relationship characteristics, to construct a 360° user profile. By integrating this multi-dimensional data, the system can more accurately identify the first and second types of users, avoiding misjudgments caused by a single indicator.

[0060] When training a credit card delinquency user classification model, with user authorization, multidimensional data of authorized users can be obtained to form a training dataset. In this training dataset, credit card users who have had delinquencies but repaid them within a certain period (which can be flexibly determined according to the delinquent amount) can be labeled as Category II users. Users who have failed to repay after a certain period and those with long-term unpaid debts registered with authoritative credit authorities can be labeled as Category I users. The model then learns the differences in multidimensional features between Category II and Category I users to achieve accurate classification.

[0061] Specifically, during data acquisition and preprocessing, a data acquisition module can be responsible for obtaining user data from multiple data sources. After obtaining the raw data, it undergoes data cleaning, deduplication, and standardization to ensure data consistency and integrity. In some embodiments, a federated learning framework can be used to achieve cross-institutional collaborative data modeling to address the data silo problem while protecting user privacy.

[0062] During the training data preprocessing stage, the SMOTE-Tomek algorithm can be used to address the sample imbalance problem, ensuring that the proportion of samples from the first type of users in the training data is not less than 30%. In addition, missing values ​​can be imputed using the K-nearest neighbor algorithm or random forest regression to fill in the missing data, thereby improving the training effect of the model.

[0063] In some embodiments, an improved XGBoost algorithm can be used to construct a credit card delinquency user classification model. This allows for the integration of the SHapley Additive exPlanations (SHAP) mechanism to quantify the impact of each feature on the classification results, enabling financial institutions to clearly understand the basis for the scoring results and thereby improving the transparency and credibility of their decisions.

[0064] Furthermore, a current difficulty in handling credit card delinquency is that the authenticity of the multi-dimensional data obtained from users (such as medical certificates, unemployment certificates, etc.) often requires manual verification to confirm whether the user faces objective difficulties. However, manual verification is not only time-consuming but also susceptible to human factors, resulting in low verification efficiency and further exacerbating the delinquency problem. Therefore, this invention can introduce a blockchain smart contract verification system to improve the authenticity of the obtained multi-dimensional user data. Specifically, a consortium blockchain network can be constructed, with nodes including banks, third-party data verification agencies, and medical institutions, to achieve distributed storage and sharing of data. Through a smart contract template library, the authenticity of materials such as medical certificates and unemployment certificates is automatically verified, and zero-knowledge proof technology is used to protect user privacy, ensuring that sensitive information is only visible to authorized nodes. Thus, after obtaining the target user's authorization, multi-dimensional user data can be obtained from this consortium blockchain network in operation S220, which is not only efficient but also ensures the data's authenticity and reliability.

[0065] Related technologies typically employ fixed installment templates, failing to adequately consider users' dynamic financial situations, such as income fluctuations and debt changes, leading to a mismatch between repayment plans and users' actual repayment capacity. Embodiments of this invention can generate personalized repayment plans through dynamic strategies. Specifically, some embodiments of this invention construct strategy optimization models based on reinforcement learning. Input parameters can include users' real-time financial data, macroeconomic indicators, and industry employment rates, thereby generating personalized repayment plans. For example, the number of installment periods (e.g., adaptive adjustment from 3 to 24 periods), tiered interest rates (e.g., base rate + risk premium), and temporary extension mechanisms can be dynamically adjusted based on changes in user income to ensure the feasibility of the repayment plan. Furthermore, Monte Carlo simulations can be used to verify the success rate of strategy execution, ensuring the stability of the plan, thereby improving collection success rates and reducing bad debt risk. A specific embodiment can be referenced. Figure 3 The illustration.

[0066] Figure 3 The illustration shows a flowchart illustrating the generation of personalized repayment plans for a second type of user in one embodiment of the present invention. Figure 3 As shown, according to an embodiment of the present invention, the above-mentioned operation S252 may include operations S301 to S303.

[0067] In operation S301, different repayment strategies are simulated within a Markov decision process framework using a strategy optimization model to obtain the cumulative reward for each strategy. The strategy optimization model is a reinforcement learning model constructed using a Markov decision process. At least one of the following is different among the different repayment strategies: the number of installments, the tiered interest rate, and the temporary deferral mechanism.

[0068] In operation S302, based on the cumulative rewards of different repayment strategies, the execution effect of different repayment strategies is obtained.

[0069] For example, user data (real-time financial, work, and health data), macroeconomic indicators, industry employment rates, and user verification materials can be used as input, represented as X1, X2, X3, ..., Xn. A reinforcement learning model can be trained using the Actor-Critic algorithm, taking the indicator data X1, X2, X3, ..., Xn as input and outputting the action probability distribution Pt, including a Softmax distribution of the installment repayment period and a Gaussian distribution of the interest rate. The model then outputs a state value estimate. A cumulative reward can be obtained based on the accumulation of these value estimates.

[0070] In operation S303, the repayment strategy with the best execution effect is selected from different repayment strategies as the personalized repayment plan for the target user.

[0071] In one embodiment, cumulative rewards can be used as the effect. In this way, operation S303 selects the repayment strategy with the highest cumulative reward as the personalized repayment plan for the target user.

[0072] In another embodiment, the success rate of different repayment strategies can be verified through Monte Carlo simulation. Then, based on the cumulative reward and success rate of each repayment strategy, the performance effect of each strategy is determined. For example, the strategy with the optimal performance can be selected by weighting the cumulative reward and success rate. Alternatively, the repayment strategy with the highest cumulative reward can be selected if the success rate meets a success rate threshold. Verifying the success rate of strategies through Monte Carlo simulation ensures the stability of the selected scheme, thereby improving the collection success rate and reducing the risk of bad debts.

[0073] Personalized repayment plans include flexible installment periods (adaptively adjustable from 3 to 24 months), tiered interest rates (base rate + risk premium), and temporary deferral mechanisms. For example, for users with fluctuating income, the installment repayment period can be dynamically adjusted to reduce the repayment pressure each month. For users with high credit scores, lower interest rates can be offered to encourage timely repayment.

[0074] Furthermore, existing systems lag behind in risk monitoring. Traditional overdue processing systems typically employ static risk assessment models, lacking the ability to monitor changes in user behavior in real time. For example, some users may initially demonstrate a strong willingness to repay, but over time, due to deteriorating economic conditions or other factors, they gradually transform into Category 1 users. However, existing systems often fail to identify this risk transformation in a timely manner, leading to an increase in bad debt rates. In this embodiment of the invention, the risk of Category 2 users transforming into Category 1 users can be monitored, enabling timely identification and intervention of user type transformation behavior.

[0075] Figure 4 The flowchart illustrating the risk of a second type of user converting to a first type of user is shown in another embodiment of the present invention.

[0076] like Figure 4 As shown, according to another embodiment of the present invention, operation S252 may include operations S401 to S405.

[0077] In operation S401, in response to a preset period or preset event, the multi-dimensional data of the target user is updated.

[0078] In operation S402, in response to the update of the multi-dimensional data of the target user, the current time series data of the target user is constructed.

[0079] In operation S403, the time series data of the target user is input into the trained real-time risk warning model to obtain the user type conversion probability output by the real-time risk warning model, wherein the real-time risk warning model is a time series prediction model built based on a long short-term memory network.

[0080] In operation S404, determine whether the user type conversion probability is higher than a preset probability threshold. If yes, proceed to operation S405; otherwise, return to operation S401 to continue monitoring.

[0081] In operation S405, when the user type conversion probability is higher than a preset probability threshold, an early warning is triggered that the target user is converted into a first type of user.

[0082] The specific implementation process of the real-time risk warning model is briefly described below.

[0083] (1) Data preparation, including users' historical transaction amount, transaction frequency, repayment amount, repayment timeliness, balance, credit limit utilization rate, etc. A time window t can be defined. If a user is overdue within this window, he / she is marked as a first-class user; otherwise, he / she is marked as a second-class user. Each user's data constitutes a time series. Each time point includes multiple features, thus organizing the data into a 3D tensor, with the shape represented as [number of samples, number of time steps, number of features].

[0084] (2) Construct a long short-term memory network model. A multi-layer long short-term memory network can be used to capture long-term dependencies in the time series. A dropout layer is added after the long short-term memory network layer to prevent overfitting. A fully connected layer can be used to output the probability of the second type of user transforming into the first type of user. Since the positive and negative samples may be unbalanced (the first type of user is less), a weighted cross-entropy loss function can be used or the minority of samples can be oversampled.

[0085] (3) Model prediction: Input the current time series data of each user into the model to obtain the probability that the user may be converted to the first type of user at a certain time in the future. Set a threshold a=0.5. If it is higher than the threshold, it is judged that the user type conversion risk is high.

[0086] The embodiments of the present invention can employ a time-series prediction model based on a long short-term memory network to monitor sudden changes in user behavior patterns, such as dramatic changes in consumption structure or social network disruptions, in order to identify potential risks of user type conversion.

[0087] In one embodiment, when the user type conversion probability is higher than a preset probability threshold, multiple quantitative indicators (such as repayment willingness decay coefficient, fund misappropriation probability, and at least one of multiple borrowing correlation) are extracted based on the target user's multi-dimensional data. When any of the quantitative indicators exceeds the corresponding indicator threshold, the target user is determined to be converted into a first-type user. The indicator thresholds corresponding to different quantitative indicators can be different. For example, when the user's repayment willingness decay coefficient exceeds the threshold, the repayment plan can be automatically adjusted, or a manual review process can be triggered to further confirm the user's true situation.

[0088] The repayment willingness decay coefficient can be determined based on whether a user adheres to installment repayment deadlines, repays in full, or experiences repayment delays and their severity across multiple repayment periods. The probability of fund misappropriation is an indicator compiled through analysis of user transaction records or transaction partners, such as the presence of unnecessary high-end consumption, to measure whether a user has the funds but refuses to repay. The multiple borrowing correlation can be extracted based on user transfer behavior and transfer recipients to determine whether a user is lending funds to others (e.g., lending activities). This invention, when detecting a high risk of a second-type user potentially converting to a first-type user, can use quantitative indicators to specifically analyze the content and causes of the risk, facilitating targeted strategies by financial institutions.

[0089] This invention utilizes a user type conversion early warning indicator system (such as key indicators like the repayment willingness decay coefficient, the probability of fund misappropriation, and the correlation between multiple borrowing) to quantify a user's risk level when a user type conversion warning is issued. By establishing a dynamic threshold mechanism, different quantitative indicators can have different thresholds. When a corresponding indicator exceeds a preset threshold, it can automatically trigger a strategy adjustment or manual review process, thereby achieving real-time monitoring and early warning of risk through fine-grained analysis.

[0090] In addition, in some embodiments, user financial data can be updated daily, and the credit card delinquency user classification model can be iterated through an online learning mechanism to adapt to changes in the market environment and user behavior.

[0091] Figure 5 A flowchart illustrating a method for handling credit card delinquency according to another embodiment of the present invention is shown.

[0092] like Figure 5 As shown, method 500 may include steps S1 to S5.

[0093] S1, the smart contract verifies and stores the data.

[0094] A blockchain network (such as a consortium blockchain) can be built, with nodes including banks, third-party data verification agencies, and medical institutions, to achieve distributed data storage and sharing. Through a smart contract template library, the authenticity of user-submitted medical certificates, unemployment certificates, and other materials is automatically verified, and zero-knowledge proof technology is used to protect user privacy, ensuring that sensitive information is only visible to authorized nodes.

[0095] After a user submits an explanation for the overdue payment and supporting documents, key information can be extracted using Optical Character Recognition (OCR) and semantic analysis technologies. The smart contract can then automatically compare the information with trusted data stored on the blockchain network to verify the authenticity of the documents. If the verification is successful, the information is stored on the blockchain network.

[0096] Specifically, the smart contract execution process includes the following steps:

[0097] (1) Users submit overdue explanations and supporting documents: Users submit overdue explanations and upload relevant supporting documents, such as medical diagnosis certificates, resignation certificates, etc., through the system interface or mobile application.

[0098] (2) OCR and semantic analysis: OCR technology can be used to extract key information from materials and combined with semantic analysis technology to verify the integrity of materials and ensure the accuracy and consistency of information.

[0099] (3) Blockchain smart contract verification: Blockchain smart contracts can be invoked to compare trusted data stored on the chain to verify the authenticity of materials. For example, medical certificates need to be compared with the blockchain records stored by medical institutions to confirm their validity.

[0100] S2, Data Acquisition and Preprocessing.

[0101] When a user's credit card delinquency data meets the criteria for credit card delinquency detection, and provided the user has authorized the use of the relevant data and possesses the corresponding permissions, the data acquisition module can retrieve the user data from the blockchain network. The retrieved raw data can then undergo data cleaning, deduplication, and standardization processes to ensure data consistency and integrity.

[0102] S3, User Classification.

[0103] The feature engineering module can be used to perform feature engineering on the preprocessed user data to construct multidimensional features of the user. Then, the credit card delinquency user classification model can be used to classify the user based on these multidimensional features to identify whether the user is a first-class user or a second-class user.

[0104] S4, dynamic repayment strategy generation and execution.

[0105] For the second type of user, a personalized repayment plan can be generated through a strategy optimization model based on reinforcement learning within the dynamic repayment strategy generation module. The model's input parameters include the user's real-time financial data, macroeconomic indicators, and industry employment rates. Through reinforcement learning algorithms, the model can simulate the execution effects of different repayment strategies and select the optimal plan.

[0106] The specific execution process of the dynamic repayment strategy generation module is briefly described below.

[0107] (1) User financial data update: User data, such as income flow, debt ratio, asset status, etc., can be updated regularly (e.g. daily) to ensure the timeliness and adaptability of repayment strategies.

[0108] (2) Reinforcement learning model training: Based on reinforcement learning algorithms, the execution effects of different repayment strategies can be simulated, and the optimal solution can be selected. The input parameters of the model include the user's real-time financial data, macroeconomic indicators, and industry employment rates.

[0109] (3) Personalized repayment plan generation: Based on the output of the reinforcement learning model, a personalized repayment plan can be generated, including flexible installment periods, tiered interest rates and temporary extension mechanisms.

[0110] (4) Monte Carlo simulation verification: The success rate of the strategy can be verified through Monte Carlo simulation to ensure the stability of the plan, thereby improving the collection success rate and reducing the risk of bad debts.

[0111] (5) Smart contract execution of repayment plan: Personalized repayment plans can be uploaded to the blockchain network and automatically executed through smart contracts to ensure the accuracy and security of fund transfers.

[0112] S5, a real-time risk monitoring and early warning mechanism.

[0113] To achieve real-time monitoring of user type conversion behavior, a real-time risk warning model based on long short-term memory networks can be used to monitor sudden changes in user behavior patterns, such as drastic changes in consumption structure or social network disruptions. Furthermore, this can be combined with a user type conversion warning indicator system (such as key indicators like the repayment willingness decay coefficient, the probability of fund misappropriation, and the correlation between multiple borrowing) to quantify the user's risk level.

[0114] The specific process of the real-time risk monitoring and early warning mechanism is briefly described below.

[0115] (1) User behavior pattern monitoring: A real-time risk warning model is used to monitor sudden changes in user behavior patterns, such as drastic changes in consumption structure and social network disruptions, in order to identify potential user type conversion risks.

[0116] (2) Calculation of user type conversion early warning indicators: Calculate user type conversion early warning indicators, including key indicators such as repayment willingness decay coefficient, fund misappropriation probability and multiple borrowing correlation, in order to quantify the user's risk level.

[0117] (3) Dynamic threshold mechanism triggering: A dynamic threshold mechanism can be established for different users and different quantitative indicators. When a quantitative indicator exceeds the preset threshold, the strategy adjustment or manual review process is automatically triggered.

[0118] (4) Manual review and strategy adjustment: If the system triggers the manual review process, the reviewers will further confirm the user's situation and adjust the repayment plan or take other risk control measures based on the review results.

[0119] (5) Real-time risk warning model iteration: User data can be updated regularly, and the credit scoring model can be iterated through an online learning mechanism to adapt to changes in the market environment and user behavior.

[0120] In one embodiment, a system for implementing the credit card overdue processing method of the present invention may adopt a layered architecture, including a data layer, a model layer, a contract layer, and an application layer.

[0121] The data layer is responsible for data collection, cleaning, and storage.

[0122] The model layer includes a credit card delinquency user classification model, a strategy optimization model, and a real-time risk warning model.

[0123] The contract layer can implement smart contract verification based on blockchain technology, verifying and storing multi-dimensional user data. In some embodiments, the contract layer can also be used to store the classification results output by the credit card delinquency user classification model and the corresponding strategies to the blockchain network.

[0124] The application layer provides the user interface and management backend.

[0125] The interaction process between modules is as follows: After a user submits an explanation of overdue payments and supporting documents, key information can be extracted using OCR and semantic analysis technologies. The smart contract can automatically compare this information with trusted data stored on the blockchain network to verify the authenticity of the documents. If verification is successful, the information is stored on the blockchain network. When a user's credit card overdue data meets the credit card overdue detection criteria, multi-dimensional data of the user is obtained with the user's authorization to classify the user using a credit card overdue user classification model. When a user is identified as a second-category user, the dynamic repayment strategy generation module can be invoked to generate a personalized repayment plan, which is then stored on the blockchain network. In this way, with the user's authorization, the smart contract can automatically execute bill installment payments and fund transfers within the blockchain network. Simultaneously, a real-time risk warning model continuously monitors user behavior patterns to identify potential user type conversion risks, ensuring intelligent and real-time overdue processing.

[0126] In this way, the embodiments of the present invention effectively solve the shortcomings of existing credit card overdue processing methods that determine repayment strategies based on fixed rules in terms of user classification, strategy adaptability, verification efficiency and risk monitoring. It realizes refined identification of user overdue behavior, formulation of personalized repayment strategies and real-time risk monitoring, which greatly improves the risk management capabilities of financial institutions and the user experience.

[0127] Based on the above-described method for handling credit card delinquency, this invention also provides a device for handling credit card delinquency. The following will be combined with... Figure 6 The device is described in detail.

[0128] Figure 6 A schematic block diagram of a credit card overdue processing device according to an embodiment of the present invention is shown.

[0129] like Figure 6 As shown, the credit card overdue processing device 600 (hereinafter referred to as device 600) according to an embodiment of the present invention may include: a data acquisition module 610, a feature engineering module 620, a user classification module 630 and a processing module 640.

[0130] The data acquisition module 610 is used to obtain user authorization to use the multi-dimensional data of the target user when the target user's credit card delinquency data meets the credit card delinquency detection conditions; and to obtain the multi-dimensional data of the target user after obtaining the user authorization. In one embodiment, the data acquisition module 610 can perform the operations S210 and S220 described above.

[0131] The feature engineering module 620 is used to construct multi-dimensional features of the target user based on the target user's multi-dimensional data. These multi-dimensional features include behavioral features, financial features, and social relationship features. In one embodiment, the feature engineering module 620 can perform the operation S230 described above.

[0132] The user classification module 630 is used to input the multidimensional features of the target user into a trained credit card delinquency user classification model to obtain the classification result output by the credit card delinquency user classification model. The classification result indicates whether the target user belongs to a first type of user or a second type of user. In one embodiment, the user classification module 630 can perform the operation S240 described above.

[0133] The processing module 640 is configured to process the credit card delinquency of the target user according to a first strategy when the target user belongs to a first type of user; and to process the credit card delinquency of the target user according to a second strategy when the target user belongs to a second type of user, wherein the second strategy is different from the first strategy. In one embodiment, the processing module 640 may perform the operations S251 and S252 described above.

[0134] According to an embodiment of the present invention, the device 600 further includes a dynamic repayment strategy generation module. The dynamic repayment strategy generation module is used to generate a personalized repayment plan for the target user based on at least some data of the target user, macroeconomic data, and employment data of the target user's industry, using a strategy optimization model constructed based on reinforcement learning. The personalized repayment plan includes installment periods, tiered interest rates, and a temporary extension mechanism, wherein at least some data of the target user comes from the target user's multi-dimensional data. In one embodiment, the dynamic repayment strategy generation module can execute operations S301 to S303 as described above.

[0135] According to an embodiment of the present invention, the device 600 further includes a real-time early warning module. The real-time early warning module is configured to: update the multi-dimensional data of the target user in response to a preset period or a preset event; construct the current time-series data of the target user in response to the update of the multi-dimensional data of the target user; input the time-series data of the target user into a trained real-time risk early warning model to obtain the user type conversion probability output by the real-time risk early warning model, wherein the real-time risk early warning model is a time-series prediction model constructed based on a long short-term memory network; and trigger an early warning that the target user has been converted into a first-type user when the user type conversion probability is higher than a preset probability threshold. In one embodiment, the real-time early warning module can perform operations S401 to S405 as described above.

[0136] Device 600 can perform reference Figures 2-5 The methods for repaying overdue credit card payments are described above and will not be repeated here.

[0137] According to embodiments of the present invention, any multiple modules among the data acquisition module 610, feature engineering module 620, user classification module 630, processing module 640, dynamic repayment strategy generation module, and real-time early warning module can be combined into one module, or any one of these modules can be split into multiple modules. Alternatively, at least some of the functions of one or more of these modules can be combined with at least some of the functions of other modules and implemented in one module. According to embodiments of the present invention, at least one of the data acquisition module 610, feature engineering module 620, user classification module 630, processing module 640, dynamic repayment strategy generation module, and real-time early warning module can be at least partially implemented as hardware circuitry, such as field-programmable gate array (FPGA), programmable logic array (PLA), system-on-a-chip, system-on-a-substrate, system-on-package, application-specific integrated circuit (ASIC), or any other reasonable means of integrating or packaging circuitry, or implemented in hardware or firmware, or in any one of software, hardware, and firmware implementations, or in a suitable combination of any of these. Alternatively, at least one of the data acquisition module 610, feature engineering module 620, user classification module 630, processing module 640, dynamic repayment strategy generation module, and real-time early warning module can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0138] Figure 7 A block diagram schematically illustrates an electronic device suitable for implementing a credit card delinquency processing method according to an embodiment of the present invention.

[0139] like Figure 7 As shown, an electronic device 900 according to an embodiment of the present invention includes a processor 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage portion 908 into a random access memory (RAM) 903. The processor 901 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 901 may also include onboard memory for caching purposes. The processor 901 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0140] RAM 903 stores various programs and data required for the operation of electronic device 900. Processor 901, ROM 902, and RAM 903 are interconnected via bus 904. Processor 901 executes various operations of the method flow according to embodiments of the present invention by executing programs in ROM 902 and / or RAM 903. It should be noted that the programs may also be stored in one or more memories other than ROM 902 and RAM 903. Processor 901 may also execute various operations of the method flow according to embodiments of the present invention by executing programs stored in said one or more memories.

[0141] According to an embodiment of the present invention, the electronic device 900 may further include an input / output (I / O) interface 905, which is also connected to a bus 904. The electronic device 900 may also include one or more of the following components connected to the input / output (I / O) interface 905: an input section 906 including a keyboard, mouse, etc.; an output section 907 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the input / output (I / O) interface 905 as needed. A removable medium 911, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 910 as needed so that computer programs read from it can be installed into the storage section 908 as needed.

[0142] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of the present invention.

[0143] According to embodiments of the present invention, a computer-readable storage medium may be a non-volatile computer-readable storage medium, such as including, but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of the present invention, a computer-readable storage medium may include ROM 902 and / or RAM 903 and / or one or more memories other than ROM 902 and RAM 903 described above.

[0144] Embodiments of the present invention also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to cause the computer system to implement the credit card overdue processing method provided in the embodiments of the present invention.

[0145] When the computer program is executed by the processor 901, it performs the functions defined in the system / apparatus of this invention. According to embodiments of the invention, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0146] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 909, and / or installed from a removable medium 911. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0147] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 909, and / or installed from the removable medium 911. When the computer program is executed by the processor 901, it performs the functions defined in the system of this embodiment of the invention. According to embodiments of the invention, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0148] According to embodiments of the present invention, program code for executing the computer programs provided in the embodiments of the present invention can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0149] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0150] Those skilled in the art will understand that the features described in the various embodiments of the present invention can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present invention. In particular, the features described in the various embodiments of the present invention can be combined and / or combined in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or combinations fall within the scope of the present invention.

Claims

1. A method for handling credit card delinquency, comprising: When the credit card delinquency data of a target user meets the credit card delinquency detection conditions, obtain user authorization to use the multi-dimensional data of the target user. With the user's authorization obtained, multi-dimensional data of the target user is acquired; Based on the multi-dimensional data of the target user, a multi-dimensional feature of the target user is constructed, which includes behavioral features, financial features, and social relationship features. The multidimensional features of the target user are input into the trained credit card delinquency user classification model to obtain the classification result output by the credit card delinquency user classification model. The classification result indicates whether the target user belongs to the first type of user or the second type of user. as well as When the target user belongs to the first type of user, the target user's credit card delinquency behavior shall be handled in accordance with the first strategy; And when the target user belongs to the second type of user, the target user's credit card delinquency behavior is handled according to the second strategy, wherein the second strategy is different from the first strategy.

2. The method according to claim 1, wherein, When the target user belongs to the second type of user, the second strategy for handling the target user's credit card delinquency includes: By using a policy optimization model built on reinforcement learning, a personalized repayment plan is generated for the target user based on at least some of the target user's data, macroeconomic data, and employment data of the target user's industry. The personalized repayment plan includes installment periods, tiered interest rates, and temporary extension mechanisms. At least some of the target user's data comes from the target user's multi-dimensional data.

3. The method according to claim 2, wherein, The strategy optimization model is a reinforcement learning model constructed using a Markov decision process, and the generation of the personalized repayment plan for the target user includes: The strategy optimization model simulates different repayment strategies within a Markov decision process framework to obtain the cumulative rewards of different repayment strategies; wherein, at least one of the following is different among the different repayment strategies: number of installments, tiered interest rate, and temporary extension mechanism; Based on the cumulative rewards of different repayment strategies, the execution effects of different repayment strategies are obtained; and The repayment strategy with the best performance among different repayment strategies is selected as the personalized repayment plan for the target user.

4. The method according to claim 3, wherein, The cumulative rewards based on different repayment strategies result in the following effects from the different repayment strategies: The success rate of different repayment strategies was verified through Monte Carlo simulations; and The execution effect of each repayment strategy is determined based on the cumulative reward and the execution success rate of each repayment strategy.

5. The method according to claim 1 or 2, wherein, The step of handling the target user's credit card delinquency behavior according to the second strategy when the target user belongs to the second type of user also includes: In response to a preset period or preset event, update the multi-dimensional data of the target user; In response to updates to the multi-dimensional data of the target user, construct the current time-series data of the target user; The time series data of the target user is input into the trained real-time risk warning model to obtain the user type conversion probability output by the real-time risk warning model, wherein the real-time risk warning model is a time series prediction model built based on a long short-term memory network. When the conversion probability of the user type is higher than the preset probability threshold, an early warning is triggered to convert the target user into the first type of user.

6. The method according to claim 5, wherein, The method further includes: When the conversion probability of the user type is higher than a preset probability threshold, multiple quantitative indicators are extracted based on the multi-dimensional data of the target user. When any of the quantitative indicators exceeds the corresponding threshold, the target user is determined to be converted into a first-type user; wherein the thresholds corresponding to different quantitative indicators are different.

7. The method according to claim 1, wherein, The acquisition of multi-dimensional data of the target user includes: acquiring multi-dimensional data of the target user from a blockchain network.

8. A device for processing overdue credit card payments, comprising: The data acquisition module is used to obtain user authorization to use the multi-dimensional data of the target user when the target user's credit card delinquency data meets the credit card delinquency detection conditions; And, with the user's authorization obtained, to acquire multi-dimensional data of the target user; The feature engineering module is used to construct multi-dimensional features of the target user based on the target user's multi-dimensional data. The multi-dimensional features include behavioral features, financial features, and social relationship features. The user classification module is used to input the multidimensional features of the target user into the trained credit card delinquency user classification model to obtain the classification result output by the credit card delinquency user classification model. The classification result indicates that the target user belongs to the first type of user or the second type of user. The processing module is used to process the target user's credit card delinquency behavior according to the first strategy when the target user belongs to the first type of user; And when the target user belongs to the second type of user, the target user's credit card delinquency behavior is handled according to the second strategy, wherein the second strategy is different from the first strategy.

9. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program or instructions stored thereon, wherein, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 7.

11. A computer program product comprising a computer program or instructions, wherein, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 7.

Images