Account processing method and device, equipment, medium and program product

Abstract

The invention provides an account processing method which can be applied to the field of big data. The account processing method comprises the steps of obtaining a plurality of risk accounts stored in a target database, determining an attribution subject of each risk account, and obtaining an attribution subject set; the risk account is an account with abnormal transaction; obtaining transaction behavior data of each attribution subject in the attribution subject set; according to the transaction behavior data, determining a risk threshold value of an affiliation subject set; the transaction behavior data and the risk threshold value of each affiliation subject are sent to a second processor; and in response to feedback information received from the second processor, each risk account is processed according to the feedback information, and the feedback information is a processing strategy for each attribution subject obtained by processing the transaction behavior data and the risk threshold value of each attribution subject by the second processor based on a pre-trained processing strategy model. The invention further provides an account processing device and equipment, a medium and a program product.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of big data and financial technology, and in particular to an account processing method and device, equipment, medium and program product. BACKGROUND

[0002] In the field of financial risk management, effectively identifying and disposing of accounts with abnormal transaction behaviors is a key link to ensure fund safety and maintain transaction order. In some examples, a common risk account processing method is based on a simple rule judgment of a single threshold, such as dividing and taking a unified disposal action according to only the account overdue duration. The processing logic of the above method is relatively simple and is difficult to cope with complex scenarios of cross-influence of multi-dimensional risk characteristics, resulting in insufficient disposal precision.

[0003] Therefore, the risk account processing method of the related art has insufficient precision and applicability in coping with large-scale and differentiated risk disposal needs, affecting the risk control effect. SUMMARY

[0004] In view of the above problems, the present application provides an account processing method and device, equipment, medium and program product for improving account processing precision.

[0005] According to a first aspect of the present application, an account processing method is provided, comprising: obtaining a plurality of risk accounts stored in a target database, determining the attribution subject of each risk account to obtain an attribution subject set; the risk account is an account with abnormal transaction; obtaining transaction behavior data of each attribution subject in the attribution subject set; determining a risk threshold of the attribution subject set according to the transaction behavior data; sending the transaction behavior data and the risk threshold of each attribution subject to a second processor; in response to receiving feedback information from the second processor, processing each risk account according to the feedback information, wherein the feedback information is a processing strategy for each attribution subject obtained by the second processor based on a pre-trained processing strategy model processing the transaction behavior data and the risk threshold of each attribution subject.

[0006] According to an embodiment of the present application, obtaining transaction behavior data of each attribution subject in the attribution subject set comprises: for any target attribution subject in the attribution subject set, determining at least one same-name account attributed to the target attribution subject; obtaining transaction behavior data of the at least one same-name account as the transaction behavior data of the target attribution subject.

[0007] According to an embodiment of the present application, obtaining transaction behavior data of the at least one same-name account comprises: obtaining a transaction abnormal duration of a risk account in the at least one same-name account; obtaining a transaction frequency of the at least one same-name account in a preset time period; obtaining a transaction amount of the risk account in the at least one same-name account.

[0008] According to an embodiment of the present application, the risk threshold of the attribution subject set is determined according to the transaction behavior data, including: respectively calculating the median of the transaction abnormal duration, the transaction frequency and the transaction amount of the attribution subject set, to obtain a first time threshold, a second frequency threshold and a third amount threshold; the first time threshold, the second frequency threshold and the third amount threshold are collectively determined as the risk threshold of the attribution subject set.

[0009] According to an embodiment of the present application, the feedback information includes a first processing strategy for the first attribution subject and a second processing strategy for the second attribution subject, and the number of risk accounts attributed to the first attribution subject is greater than the number of risk accounts attributed to the second attribution subject in the plurality of risk accounts; processing each risk account according to the feedback information, including: processing the risk accounts attributed to the first attribution subject according to the first processing strategy; the first processing strategy is used to instruct to process the risk accounts by using a predetermined first execution action in a first execution channel; processing the risk accounts attributed to the second attribution subject according to the second processing strategy; the second processing strategy is used to instruct to process the risk accounts by using a predetermined second execution action in a second execution channel; wherein the first execution channel and the second execution channel are different, and the execution intensity corresponding to the first execution action is greater than the execution intensity corresponding to the second execution action.

[0010] According to an embodiment of the present application, the method further comprises: in response to not receiving feedback information from the second processor within a preset time period, processing each risk account according to a third processing strategy; wherein the third processing strategy is used to instruct to process each risk account by using a predetermined third execution action in a third execution channel, the execution intensity corresponding to the third execution action is greater than the execution intensity corresponding to the second execution action, and less than the execution intensity corresponding to the first execution action.

[0011] According to an embodiment of the present application, the method further comprises: in response to receiving evaluation information of the processing result of each risk account, sending the evaluation information to the second processor to make the second processor retrain the pre-trained processing strategy model according to the evaluation information; wherein the evaluation information includes positive evaluation information and negative evaluation information.

[0012] The second aspect of the present application provides an account processing device, comprising: a first determination module configured to obtain a plurality of risk accounts stored in a target database, determine a belonging subject of each risk account, and obtain a belonging subject set; the risk account is an account with transaction anomalies; an acquisition module configured to acquire transaction behavior data of each belonging subject in the belonging subject set; a second determination module configured to determine a risk threshold of the belonging subject set according to the transaction behavior data; a sending module configured to send the transaction behavior data of each belonging subject and the risk threshold to a second processor; and a processing module configured to, in response to receiving feedback information from the second processor, process each risk account according to the feedback information, wherein the feedback information is a processing strategy for each belonging subject obtained by the second processor processing the transaction behavior data of each belonging subject and the risk threshold based on a pre-trained processing strategy model.

[0013] The third aspect of the present application provides an electronic device, comprising: one or more processors; a memory configured to store 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.

[0014] The fourth aspect of the present application further provides a computer-readable storage medium having a computer program or instructions stored thereon, wherein the computer program or instructions are executed by a processor to implement the steps of the method.

[0015] The fifth aspect of the present application further provides a computer program product comprising a computer program or instructions, wherein the computer program or instructions are executed by a processor to implement the steps of the method. BRIEF DESCRIPTION OF DRAWINGS

[0016] The above and other objects, features and advantages of the present application will become more apparent from the following description of embodiments of the present application, taken in conjunction with the accompanying drawings, in which:

[0017] Figure 1 An application scenario diagram of an account processing method, device, equipment, medium and program product according to an embodiment of the present application is schematically shown;

[0018] Figure 2 A flowchart of an account processing method according to an embodiment of the present application is schematically shown;

[0019] Figure 3 A flowchart of an account processing method according to another embodiment of the present application is schematically shown;

[0020] Figure 4 A structural block diagram of an account processing device according to an embodiment of the present application is schematically shown; and

[0021] Figure 5A block diagram of an electronic device suitable for implementing the account processing method according to embodiments of the present application is shown schematically. DETAILED DESCRIPTION

[0022] Embodiments of the present application will be described below with reference to the accompanying drawings. It should be understood, however, that the description that follows is merely exemplary and is not intended to limit the scope of the application. In the following detailed description of embodiments of the present application, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one skilled in the art that one or more embodiments of the present application can be practiced without these specific details. In other instances, well-known structures and functions have not been described in detail in order to avoid obscuring aspects of the present application.

[0023] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the present application. As used herein, the term "includes" and tautological expressions thereof, such as "including," "includes," "include," "contains," "containing," and so forth, shall not be taken to exclude

[0024] All terms used herein including technical and scientific terms have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined herein. It should be noted that the terms used herein are defined as consistent with the meaning that is consistent with the context of the specification unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of the specification, and should not be interpreted in an idealized or overly formal manner.

[0025] In the case of using expressions similar to "at least one of A, B, and C, etc.", it should be generally interpreted that the meaning of the expression is the same as that of "at least one of A and B; at least one of B and C; at least one of A and C; at least one of A, B, and C, etc.".

[0026] In the technical solutions of the present application, the user information (including but not limited to user personal information, user image information, user device information, such as location information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved are information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards, necessary security measures are taken, do not violate public order and good customs, and provide corresponding operation portal for user to choose authorization or refusal.

[0027] In the scenario of making automated decisions by using personal information, the method, device and system provided by the embodiments of the present application all provide corresponding operation entrances for the user to select to agree or reject the automated decision result; if the user selects to reject, the expert decision process is entered. The expression "automated decision" here refers to the activity of automatically analyzing, evaluating the personal behavior habits, interests and hobbies, or economic, health, credit status, etc. by a computer program, and making decisions. The expression "expert decision" here refers to the activity of making decisions by personnel who are engaged in a certain field of work, have special experience, knowledge and skills, and reach a certain professional level.

[0028] Figure 1 An application scenario diagram of the account processing method, apparatus, device, medium and program product according to the embodiments of the present application is schematically shown.

[0029] As Figure 1 shown, the application scenario 100 according to the embodiments can include the field of financial technology. The network 104 is a medium for providing communication links between the first terminal device 101, the second terminal device 102, the third terminal device 103 and the server 105. The network 104 can include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0030] The user can use the first terminal device 101, the second terminal device 102, the third terminal device 103 to interact with the server 105 through 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, the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only as examples).

[0031] The first terminal device 101, the second terminal device 102, the third terminal device 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers and desktop computers, etc.

[0032] The server 105 can be a server providing various services, such as a background management server supporting the website browsed by the user using the first terminal device 101, the second terminal device 102, the third terminal device 103 (only as an example). The background management server can analyze and process the received user request data, etc., and feed back the processing result (such as a web page, information, or data, etc. obtained or generated according to the user request) to the terminal device.

[0033] It should be noted that the account processing method provided in the embodiments of the present application can be generally executed by the server 105. Accordingly, the account processing apparatus provided in the embodiments of the present application can be generally arranged in the server 105. The account processing method provided in the embodiments of the present application can also be executed by a server or a server cluster different from the server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Accordingly, the account processing apparatus provided in the embodiments of the present application can also be arranged in a server or a server cluster different from the server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105.

[0034] It should be understood that Figure 1 The number of terminal devices, networks and servers in the above-mentioned scenario is only illustrative. Any number of terminal devices, networks and servers can be provided according to implementation needs.

[0035] The account processing method according to the embodiments of the present application will be described in detail below based on the scenario described above. Figure 1 Figures 2-5 The account processing method according to the embodiments of the present application will be described in detail below based on the scenario described above.

[0036] Figure 2 An illustrative flowchart of the account processing method according to the embodiments of the present application is shown.

[0037] As shown in Figure 2 The account processing method of this embodiment includes operations S210-S250, which can be executed by a server.

[0038] In operation S210, a plurality of risk accounts stored in a target database are obtained, and a belonging subject of each risk account is determined to obtain a belonging subject set; the risk account is an account with transaction anomaly.

[0039] In the embodiments of the present application, the risk account refers to an account identified as having transaction anomaly in a financial or transaction system.

[0040] In the embodiments of the present application, the belonging subject refers to a legal or business responsibility bearer of the risk account, for example, the belonging subject can be an individual or institutional customer. The belonging subject set refers to a set composed of the belonging subjects corresponding to all identified risk accounts after deduplication.

[0041] ​In embodiments of the present application, a plurality of risk accounts stored in a target database are obtained, and a belonging subject of each risk account is determined to obtain a belonging subject set. The first processor accesses a target database (such as a core transaction database or a risk monitoring database), obtains a list of all accounts currently marked as “having transaction abnormalities” through query or receiving an alarm, that is, a plurality of risk accounts. According to the account information, the actual controller or responsible party behind each account is determined, that is, the belonging subject, and is summarized to form a list of independent customer groups that need to be processed, that is, the belonging subject set.

[0042] For example, the first processor obtains a batch of bank card numbers triggering suspicious transaction warnings from the risk monitoring database, queries the corresponding cardholder information according to the card numbers, and forms a customer list to be processed.

[0043] In operation S220, transaction behavior data of each belonging subject in the belonging subject set is obtained.

[0044] In embodiments of the present application, the transaction behavior data refers to a data index used to quantitatively describe and evaluate the risk status, behavior pattern or financial status of a belonging subject.

[0045] In embodiments of the present application, transaction behavior data of each belonging subject in the belonging subject set is obtained. For each belonging subject, the first processor extracts or calculates dynamic indexes that can reflect the current risk level from its associated data sources. These data may be scattered in transaction flow, account archives, behavior logs and other databases.

[0046] For example, for a cardholder listed as a risk customer, the system may obtain “the number of days from the last suspicious transaction to today”, “the total number of transactions in the past month”, “the total balance of associated accounts” and the like as transaction behavior data.

[0047] In operation S230, a risk threshold of the belonging subject set is determined according to the transaction behavior data.

[0048] In embodiments of the present application, the risk threshold refers to one or more reference benchmark values calculated by a statistical method based on the distribution of a certain transaction behavior data of the entire belonging subject set. The risk threshold is used to establish a division standard at the group level.

[0049] In embodiments of the present application, a risk threshold of the belonging subject set is determined according to the transaction behavior data. The first processor analyzes the data distribution of the belonging subject set in a certain dimension (for example, “the number of days from the last suspicious transaction” of all customers), and determines the risk threshold of the belonging subject set by using a specific statistical algorithm (such as calculating the median, average or specific quantile).

[0050] For example, the 75% quantile of the number of transactions of all risky customers in the past month is calculated, and the number of transactions is used as the threshold of high-frequency trading.

[0051] In operation S240, the transaction behavior data of each belonging subject and the risk threshold are sent to the second processor.

[0052] In the embodiments of the present application, the second processor refers to an independent processing unit or system that communicates with the first processor and has model calculation capability.

[0053] In the embodiments of the present application, the transaction behavior data of each belonging subject and the risk threshold are sent to the second processor. The first processor packages the prepared individual data (the transaction behavior data of each belonging subject) and the group reference (the risk threshold), and sends them to the second processor responsible for intelligent decision-making through a network interface or a message queue.

[0054] In operation S250, in response to receiving feedback information from the second processor, each risk account is processed according to the feedback information, wherein the feedback information is a processing strategy for each belonging subject obtained by the second processor based on the pre-trained processing strategy model processing the transaction behavior data of each belonging subject and the risk threshold.

[0055] In the embodiments of the present application, the feedback information refers to the information returned by the second processor to the first processor after receiving the data, and the core content is the processing strategy for each belonging subject. The processing strategy refers to specific, executable operation instructions or action plans that should be taken for the risk account or the belonging subject.

[0056] In the embodiments of the present application, in response to receiving feedback information from the second processor, each risk account is processed according to the feedback information. The first processor listens to and receives feedback from the second processor. The second processor internally uses its pre-trained processing strategy model to analyze, evaluate or compare the received data and threshold of each customer (belonging subject), generates a customized optimal or recommended processing strategy for each customer, and sends it back as feedback information. After receiving the feedback information, the first processor parses the feedback information and executes specific processing operations on each original risk account according to the strategy indicated therein.

[0057] For example, the feedback information indicates that the strategy of "raising the monitoring level and sending an alert SMS" is adopted for customer A, and the first processor calls the monitoring system and the SMS platform to execute this operation. It should be noted that the processing strategy is formulated for the customer (belonging subject), and the execution of the processing strategy finally lands on the risk account under the name of the customer.

[0058] Through the embodiments of the present application, the first processor is responsible for efficiently and reliably completing pre-work such as customer identification, data aggregation, and threshold calculation, ensuring the quality and consistency of the input data; the second processor focuses on using a pre-trained model to make complex strategy decisions, improving the intelligent level and precision of processing. The method of the present embodiment significantly improves the efficiency and system reliability of large-scale risk account processing, and improves the accuracy and reliability of risk management.

[0059] Figure 3 A flowchart of an account processing method according to another embodiment of the present application is schematically shown.

[0060] As shown in Figure 3 In the embodiments of the present application, the transaction behavior data of each attribution subject in the attribution subject set is obtained, including operations S310-S330, and the account processing method can be executed by a server.

[0061] In operation S310, for any target attribution subject in the attribution subject set, at least one same-name account belonging to the target attribution subject is determined.

[0062] In the embodiments of the present application, the target attribution subject refers to a specific customer currently being processed, selected from the attribution subject set. For example, customer A is selected as the object of this processing from the list of customers to be processed.

[0063] In the embodiments of the present application, the same-name account refers to all accounts in the banking system that are completely bound to the identity information (such as the ID number) of the target attribution subject. These same-name accounts belong to the same customer, for example, multiple credit cards, savings cards, etc. under the name of customer A.

[0064] In the embodiments of the present application, the transaction behavior data refers to the original data obtained from the account level, which can reflect the account activity and risk situation.

[0065] In the embodiments of the present application, for any target attribution subject in the attribution subject set, at least one same-name account belonging to the target attribution subject is determined. The system obtains the list of all accounts opened by the current customer (target attribution subject) in the bank by querying the customer information database.

[0066] For example, when processing customer A, the system queries a credit card account and a current deposit account under his name through his ID number.

[0067] In operation S320, the transaction behavior data of the at least one same-name account is obtained and used as the transaction behavior data of the target attribution subject.

[0068] In the embodiments of the present application, the system obtains the transaction behavior data of all the same-name accounts (including risk accounts and normal accounts) of the customer, and aggregates the account-level data as transaction behavior data representing the overall status of the customer (target attribution subject).

[0069] Through the embodiments of the present application, by obtaining the data of all the accounts under the customer and aggregating them into customer-level indicators, the comprehensiveness and accuracy of the customer risk portrait are ensured, and the situation of incorrect judgment caused by one-sided single account data is avoided.

[0070] In some embodiments, obtaining the transaction behavior data of at least one same-name account includes: obtaining the transaction abnormal duration of the risk account in at least one same-name account; obtaining the transaction frequency of at least one same-name account within a preset time period; obtaining the transaction amount of the risk account in at least one same-name account.

[0071] In the embodiments of the present application, the transaction abnormal duration refers to the number of days from the first abnormal state of the risk account to the current time. For example, a fund transaction has been overdue for 45 days from the first overdue date to the current date.

[0072] In the embodiments of the present application, the transaction frequency within a preset time period refers to the total number of valid transactions of the same-name account within a certain pre-set time period. For example, the total number of transactions such as offline consumption and online payment of all bank cards of the customer within the last 30 days is 20 times.

[0073] In the embodiments of the present application, the transaction amount refers to the total amount of the abnormal transactions involved in the risk account.

[0074] In the embodiments of the present application, the transaction abnormal duration of the risk account in at least one same-name account is obtained. The system calculates the duration of the abnormal state from the contract record, transaction plan and other data of the risk account under the target attribution subject. The transaction abnormal duration is used to evaluate the new and old degree and urgency of the risk state.

[0075] In the embodiments of the present application, the transaction frequency of at least one same-name account within a preset time period is obtained. The system scans the transaction flow of all same-name accounts of the target attribution subject within a preset time period (such as the last 30 days), and counts the total number of all successful transactions. The transaction frequency is used to evaluate the overall activity of the customer and the interaction frequency with the bank.

[0076] In the embodiments of the present application, the transaction amount of the risk account in at least one same-name account is obtained. The system extracts the total amount of funds directly related to abnormal behavior from the current state of the risk account. The transaction amount is used to quantify the scale of the fund loss caused by the risk event.

[0077] Through the embodiments of the present application, by precisely defining the core data of three dimensions, the abstract risk assessment is converted into quantitative analysis of specific indicators such as duration, active frequency, and fund size, which provides objective data support for subsequent accurate risk grading and differentiated strategy making, and effectively improves the accuracy of risk identification.

[0078] In some embodiments, according to the transaction behavior data, the risk threshold of the attribution subject set is determined, including: respectively calculating the median of the transaction abnormal duration, the transaction frequency and the transaction amount of the attribution subject set, to obtain a first time threshold, a second number threshold and a third amount threshold; and the first time threshold, the second number threshold and the third amount threshold are collectively determined as the risk threshold of the attribution subject set.

[0079] In the embodiments of the present application, the first time threshold refers to the median calculated based on the transaction abnormal duration of the attribution subject set, the second number threshold refers to the median calculated based on the transaction frequency of the attribution subject set, and the third amount threshold refers to the median calculated based on the transaction amount of the attribution subject set. Each median is the critical value of the risk division of the dimension.

[0080] In the embodiments of the present application, the system processes three types of data of the attribution subject set respectively, and for each type of data, the system sorts the index values of all customers and finds the median. For example, the median of the transaction abnormal duration of the attribution subject set is calculated to obtain the first time threshold (such as 60 days). The median of the transaction frequency of the attribution subject set is calculated to obtain the second number threshold (such as 8 times). The median of the transaction amount of the attribution subject set is calculated to obtain the third amount threshold (such as 50,000 yuan).

[0081] In the embodiments of the present application, the first time threshold, the second number threshold and the third amount threshold are collectively determined as the risk threshold of the attribution subject set. The three independent thresholds representing the general level of each dimension are combined together and collectively used as the comprehensive risk threshold for risk analysis and stratification of the customer group.

[0082] Through the embodiments of the present application, by using the median as a statistical quantity to objectively determine the risk threshold of each dimension, the interference of extreme values on the stratification standard can be avoided, and it is ensured that the threshold can truly reflect the overall distribution center of the customer group, so that the subsequent risk grading is more stable and reasonable.

[0083] In some embodiments, the feedback information comprises a first processing strategy for a first attribution subject and a second processing strategy for a second attribution subject, the number of risk accounts attributed to the first attribution subject is greater than the number of risk accounts attributed to the second attribution subject; processing each risk account according to the feedback information comprises: processing the risk accounts attributed to the first attribution subject according to the first processing strategy; the first processing strategy is used to instruct to process the risk accounts by using a predetermined first execution action in a first execution channel; processing the risk accounts attributed to the second attribution subject according to the second processing strategy; the second processing strategy is used to instruct to process the risk accounts by using a predetermined second execution action in a second execution channel; wherein the first execution channel and the second execution channel are different, and the execution intensity corresponding to the first execution action is greater than the execution intensity corresponding to the second execution action.

[0084] In embodiments of the present application, the first processing strategy and the second processing strategy refer to the attribution subject of different risk characteristics, and the differentiated treatment scheme generated by the processing strategy model. For example, a strong intervention strategy is adopted for high-risk customers, and a standard reminder strategy is adopted for low-risk customers.

[0085] In embodiments of the present application, the execution channel refers to the technical or business path relied on to implement the processing strategy. For example, the first execution channel can be a bank internal SMS platform, and the second execution channel can be an outsourcing call center system.

[0086] In embodiments of the present application, the execution action refers to a specific operation instruction specified in the processing strategy. For example, the first execution action can be "manual telephone communication reminder", and the second execution action can be "system automatic sending of reminder SMS".

[0087] In embodiments of the present application, the execution intensity refers to the level difference of the processing strategy in resource investment and intervention intensity. For example, one-on-one manual communication belongs to high intensity, and batch SMS reminder belongs to low intensity.

[0088] In embodiments of the present application, after the first processor receives the feedback information returned by the second processor, the specific strategy for different customers is parsed. For example, for the first attribution subject with a large number of risk accounts, the first execution action (manual telephone reminder) is executed through the first execution channel (such as an outsourcing call center) according to the first processing strategy. For the second attribution subject with a small number of risk accounts, the second execution action (system automatic sending of reminders) is executed through the second execution channel (such as a bank SMS platform) according to the second processing strategy.

[0089] In the embodiments of the present application, different risk level customers are processed differently through the combination of different execution channels and execution actions. For example, high-risk customers are communicated with high-intensity phone calls, and low-risk customers are reminded with low-intensity short messages, so as to optimize the allocation of resources.

[0090] Through the embodiments of the present application, the precise allocation of disposal resources is realized by establishing the corresponding relationship between the differentiated processing strategy and the customer risk characteristics, which not only ensures the effective intervention of high-risk customers, but also avoids the excessive disturbance of low-risk customers, and significantly improves the fine level and operation efficiency of risk management.

[0091] In some embodiments, the method further comprises: in response to not receiving feedback information from the second processor within a preset time period, processing each risk account according to a third processing strategy; wherein the third processing strategy is used to instruct processing each risk account by using a predetermined third execution action in the third execution channel, and the execution strength corresponding to the third execution action is greater than the execution strength corresponding to the second execution action and less than the execution strength corresponding to the first execution action.

[0092] In the embodiments of the present application, the preset time period refers to the maximum time limit set for the first processor to wait for the feedback information returned by the second processor after sending data. For example, the preset time period can be set to 30 seconds or 60 seconds, and if it exceeds this time limit, it is considered that the communication is abnormal or the processing is timed out.

[0093] In the embodiments of the present application, the third processing strategy refers to a default or degraded processing scheme automatically enabled by the system when the intelligent decision feedback from the second processor cannot be obtained in time. The third execution channel refers to a specific technical path or system interface for implementing the third processing strategy. The third execution action refers to a specific operation instruction predefined in the third processing strategy. For example, “sending a standard reminder message to the customer through the system and recording the follow-up state”.

[0094] In the embodiments of the present application, after the first processor sends data to the second processor, a timer is started for waiting. If no feedback information is received within the preset time period (for example, within 30 seconds), it is determined that the communication is timed out or the second processor processing is abnormal. At this time, the system automatically triggers the backup processing flow and no longer waits indefinitely. The system operates each risk account that needs to be processed according to the predefined third processing strategy. The strategy executes the third execution action through the third execution channel.

[0095] For example, through the bank's batch task system, a unified reminder message is automatically sent to the owner of all risk accounts to be processed, and these accounts are marked as “need to be checked by human beings later”.

[0096] Through the embodiments of the present application, by setting a timeout judgment mechanism and a pre-device strategy, the intelligent decision system can automatically enable a degradation processing scheme when responding to an exception, effectively ensuring the continuity and system robustness of the risk account processing flow, avoiding business interruption caused by a single point failure, and ensuring the continuous operation of basic risk management functions.

[0097] In some embodiments, the second processor processes the transaction behavior data and the risk threshold of each attribution subject to obtain feedback information by the following steps: determining a first comparison result according to the transaction abnormal duration of at least one target account of each attribution subject and a first time threshold, determining a second comparison result according to the transaction frequency of the at least one target account and a second frequency threshold, and determining a third comparison result according to the transaction amount of the at least one target account and a third amount threshold; and processing the first comparison result, the second comparison result and the third comparison result corresponding to each attribution subject by using a pre-trained processing strategy model to obtain processing strategy information for each attribution subject.

[0098] In the embodiments of the present application, the first comparison result, the second comparison result and the third comparison result respectively refer to the binary quantization results obtained by comparing the values of the transaction abnormal duration, the transaction frequency and the transaction amount of the attribution subject with the first time threshold, the second frequency threshold and the third amount threshold respectively.

[0099] For example, the comparison result can be quantified as "1" (indicating that the value is greater than or equal to the threshold) or "0" (indicating that the value is less than the threshold).

[0100] In the embodiments of the present application, the pre-trained processing strategy model refers to a trained machine learning model, which can predict the corresponding strategy according to the input feature vector by learning the mapping relationship between the comparison result combination and the optimal processing strategy in the historical data. The processing strategy information refers to the specific, executable operation suggestion or instruction code output by the model for each attribution subject.

[0101] For example, the pre-trained processing strategy model can be trained based on a decision tree model, a random forest model or a gradient boosting decision tree model, etc.

[0102] In the embodiments of the present application, the second processor performs a comparison calculation for each received data of the home principal. For example, for a home principal whose transaction abnormal duration is 30 days (the first time threshold is 60 days, 30 < 60), the first comparison result is “0”. The transaction frequency thereof is 10 times (the second frequency threshold is 8 times, 10 > 8), and the second comparison result is “1”. The transaction amount thereof is 30,000 yuan (the third amount threshold is 50,000 yuan, 30,000 < 50,000), and the third comparison result is “0”. Finally, the comparison result combination of the customer is (0, 1, 0).

[0103] In the embodiments of the present application, the second processor inputs the comparison result combination (such as (0, 1, 0)) as a feature vector into a pre-trained processing strategy model. The model internally performs forward propagation calculation through the learned parameters thereof, and finally outputs the processing strategy information corresponding to the feature vector.

[0104] For example, the model outputs the strategy code “S02”, which corresponds to “sending a standard reminder SMS”.

[0105] Through the embodiments of the present application, the second processor converts continuous business data into discrete comparison results, and utilizes a pre-trained model to perform automatic strategy matching, thereby converting a complex risk assessment decision-making process into an efficient and reusable technical process, ensuring the objectivity, consistency and processing efficiency of the strategy generation.

[0106] In some embodiments, the pre-trained processing strategy model is obtained by the following steps: obtaining a historical home principal sample set, wherein the historical home principal sample set includes a plurality of groups of training data, each group of training data including historical transaction behavior data, a historical risk threshold, and a corresponding historical processing strategy; taking the historical transaction behavior data and the historical risk threshold as input features, and taking the corresponding historical processing strategy as a prediction target, performing supervised training on an initial model to obtain the pre-trained processing strategy model.

[0107] In the embodiments of the present application, the historical home principal sample set refers to a data set used for model training, which is composed of a large number of historical cases. The training data refers to a single sample in the sample set, which is a complete triple including historical transaction behavior data, a historical risk threshold, and a historical processing strategy. The historical transaction behavior data refers to the actual transaction behavior data of a home principal at a certain time point in history. The historical risk threshold is a risk threshold calculated based on the historical transaction behavior data and the historical data of the home principal group at the same period. The historical processing strategy is the processing strategy actually adopted for the home principal at that time, which is verified as effective or optimal by subsequent effects.

[0108] In embodiments of the present application, the prediction target refers to the correct output that the model needs to learn during training, which here specifically refers to the historical handling strategy corresponding to the input features. Supervised training is a machine learning training paradigm that uses a sample set containing input features and corresponding correct outputs (prediction targets) for training.

[0109] In embodiments of the present application, data is collected from historical business records to construct a training set. For example, records of 100,000 risky customers in the past year are collected, each record containing the customer's "transaction abnormal duration", "transaction frequency", "transaction amount" (historical transaction behavior data), the calculated median threshold of the total customer at that time (historical risk threshold), and the "handling strategy" (historical handling strategy) taken at that time and ultimately successful for that customer.

[0110] In embodiments of the present application, the training data is input into the training process. The historical transaction behavior data and the historical risk threshold are used as input features, and the corresponding historical handling strategy is used as the prediction target, to train the initial model. By continuously adjusting the model parameters through optimization algorithms (such as gradient descent), the gap between the model's predicted strategy and the actual historical strategy is narrowed, resulting in a pre-trained handling strategy model that can accurately predict the handling strategy based on the input features.

[0111] Through embodiments of the present application, the handling strategy model can automatically learn the decision-making rules from a large amount of historical successful experience through supervised training methods, converting the experience of business experts into reusable model parameters, ensuring the reliability and effectiveness of model decision-making.

[0112] In some embodiments, the method further comprises: in response to receiving evaluation information of the handling result for each risk account, sending the evaluation information to a second processor to cause the second processor to retrain the pre-trained handling strategy model according to the evaluation information; wherein the evaluation information includes positive evaluation information and negative evaluation information.

[0113] In embodiments of the present application, the evaluation information of the handling result refers to the quantitative or qualitative feedback data generated by evaluating the actual effect of the strategy after the handling strategy is executed. This information reflects the effectiveness of the handling strategy.

[0114] In embodiments of the present application, positive evaluation information refers to evaluation information indicating that the handling strategy achieves the expected or good effect. For example, the risk account returns to normal within a specified period after the strategy is executed, the customer completes the overdue fund settlement, etc. Negative evaluation information refers to evaluation information indicating that the handling strategy does not achieve the expected effect or has poor effect. For example, the risk account continues to deteriorate, does not respond within a predetermined observation period, etc.

[0115] In the embodiments of the present application, the first processor obtains the effect feedback, i.e., evaluation information, for the processed risk account through a business system interface or manual input, etc. The first processor packages the evaluation information together with the corresponding customer identifier and other information and sends it to the second processor. For example, the system records that after the "manual telephone reminder" strategy is performed on customer A, the customer processes the transaction exception within 3 days, and then generates a positive evaluation information and sends it.

[0116] In the embodiments of the present application, after receiving the evaluation information, the second processor associates it with the original data (transaction behavior data, risk threshold) used when generating the processing strategy for the customer previously, to form a new training sample. The pre-trained processing strategy model is retrained using these new samples containing result feedback.

[0117] For example, the model strengthens the weight of the strategy under the corresponding feature condition according to a large number of "adopt strategy X and the result is positive" samples.

[0118] Through the embodiments of the present application, by establishing an evaluation feedback closed loop of the processing result and using the evaluation information for retraining of the model, the processing strategy model has the ability of continuous learning and self-optimization, can dynamically adapt to business changes, and thus improves the accuracy and effectiveness of the processing strategy generation.

[0119] Based on the above account processing method, the present application further provides an account processing device. The following will be described in detail Figure 4 The device.

[0120] Figure 4 The structure block diagram of the account processing device according to the embodiments of the present application is schematically shown.

[0121] As Figure 4 shown, the account processing device 400 of the embodiments includes a first determination module 410, an acquisition module 420, a second determination module 430, a sending module 440 and a processing module 450.

[0122] The first determination module 410 is configured to obtain a plurality of risk accounts stored in a target database, determine the attribution subject of each risk account, and obtain an attribution subject set; the risk account is an account with transaction anomaly. In an embodiment, the first determination module 410 can be configured to perform the operation S210 described in the foregoing, and details are not described herein again.

[0123] The acquisition module 420 is configured to obtain the transaction behavior data of each attribution subject in the attribution subject set. In an embodiment, the first determination module 420 can be configured to perform the operation S220 described in the foregoing, and details are not described herein again.

[0124] The second determining module 430 is configured to determine the risk threshold of the set of attributed subjects according to the transaction behavior data. In an embodiment, the second determining module 430 can be configured to perform operation S230 described above, and details are not repeated here.

[0125] The sending module 440 is configured to send the transaction behavior data and the risk threshold of each attributed subject to the second processor. In an embodiment, the sending module 440 can be configured to perform operation S240 described above, and details are not repeated here.

[0126] The processing module 450 is configured to, in response to receiving feedback information from the second processor, process each risk account according to the feedback information, wherein the feedback information is a processing strategy for each attributed subject obtained by the second processor based on the pre-trained processing strategy model processing the transaction behavior data and the risk threshold of each attributed subject. In an embodiment, the processing module 450 can be configured to perform operation S250 described above, and details are not repeated here.

[0127] According to embodiments of the present application, any of the first determining module 410, the obtaining module 420, the second determining module 430, the sending module 440 and the processing module 450 can be combined in one module, or any of them can be split into multiple modules. Alternatively, at least part of the function of one or more of these modules can be combined with at least part of the function of other modules, and implemented in one module. According to embodiments of the present application, at least one of the first determining module 410, the obtaining module 420, the second determining module 430, the sending module 440 and the processing module 450 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on board, a system on package, an application specific integrated circuit (ASIC), or any other reasonable way of hardware or firmware that can be integrated or packaged, or any one of software, hardware and firmware or any appropriate combination of several of them. Alternatively, at least one of the first determining module 410, the obtaining module 420, the second determining module 430, the sending module 440 and the processing module 450 can be at least partially implemented as a computer program module that can perform corresponding functions when running.

[0128] In some embodiments, the obtaining module comprises: a first determining sub-module configured to determine, for any target attributed subject in the set of attributed subjects, at least one same-name account attributed to the target attributed subject; and an obtaining sub-module configured to obtain transaction behavior data of the at least one same-name account as the transaction behavior data of the target attributed subject.

[0129] In some embodiments, the obtaining sub-module comprises: a first obtaining unit, configured to obtain the transaction abnormal duration of the risk account in the at least one same-name account; a second obtaining unit, configured to obtain the transaction frequency of the at least one same-name account in a preset time period; and a third obtaining unit, configured to obtain the transaction amount of the risk account in the at least one same-name account.

[0130] In some embodiments, the second determining module comprises: a calculation sub-module, configured to respectively calculate the median of the transaction abnormal duration, the transaction frequency and the transaction amount of the attribution subject set, to obtain a first time threshold, a second frequency threshold and a third amount threshold; and a second determining sub-module, configured to collectively determine the first time threshold, the second frequency threshold and the third amount threshold as the risk threshold of the attribution subject set.

[0131] In some embodiments, the feedback information comprises a first processing strategy for the first attribution subject and a second processing strategy for the second attribution subject, the number of risk accounts attributed to the first attribution subject in the plurality of risk accounts is greater than the number of risk accounts attributed to the second attribution subject; the processing module comprises: a first processing sub-module, configured to process the risk accounts attributed to the first attribution subject according to the first processing strategy; the first processing strategy is configured to instruct to process the risk accounts by using a predetermined first execution action in a first execution channel; a second processing sub-module, configured to process the risk accounts attributed to the second attribution subject according to the second processing strategy; the second processing strategy is configured to instruct to process the risk accounts by using a predetermined second execution action in a second execution channel; wherein the first execution channel and the second execution channel are different, and the execution intensity corresponding to the first execution action is greater than the execution intensity corresponding to the second execution action.

[0132] In some embodiments, the device further comprises: a first execution module, configured to, in response to not receiving the feedback information from the second processor within the preset time period, process each risk account according to a third processing strategy; wherein the third processing strategy is configured to instruct to process each risk account by using a predetermined third execution action in a third execution channel, the execution intensity corresponding to the third execution action is greater than the execution intensity corresponding to the second execution action and less than the execution intensity corresponding to the first execution action.

[0133] In some embodiments, the device further comprises: a second execution module, configured to, in response to receiving the evaluation information of the processing result of each risk account, send the evaluation information to the second processor, so that the second processor re-trains the pre-trained processing strategy model according to the evaluation information; wherein the evaluation information comprises positive evaluation information and negative evaluation information.

[0134] Figure 5 A block diagram of an electronic device suitable for implementing the account processing method according to embodiments of the present application is schematically shown.

[0135] like Figure 5 As shown, an electronic device 500 according to an embodiment of this application includes a processor 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage portion 508 into a random access memory (RAM) 503. The processor 501 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 501 may also include onboard memory for caching purposes. The processor 501 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0136] RAM 503 stores various programs and data required for the operation of electronic device 500. Processor 501, ROM 502, and RAM 503 are interconnected via bus 504. Processor 501 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 502 and / or RAM 503. It should be noted that programs may also be stored in one or more memories other than ROM 502 and RAM 503. Processor 501 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in one or more memories.

[0137] According to embodiments of this application, the electronic device 500 may further include an input / output (I / O) interface 505, which is also connected to a bus 504. The electronic device 500 may also include one or more of the following components connected to the input / output (I / O) interface 505: an input section 506 including a keyboard, mouse, etc.; an output section 507 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the input / output (I / O) interface 505 as needed. A removable medium 511, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 510 as needed so that computer programs read from it can be installed into the storage section 508 as needed.

[0138] The application further provides a computer readable storage medium, which can be included in the device / apparatus / system described in the above embodiments, or can exist independently without being 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 application.

[0139] According to the embodiments of the application, the computer readable storage medium can be a non-volatile computer readable storage medium, which can include, but is not limited to, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any appropriate combination thereof. In this application, a computer readable storage medium can be any tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. For example, according to the embodiments of the application, the computer readable storage medium can include one or more of the above-described ROM 502 and / or RAM 503 and / or memories other than the ROM 502 and the RAM 503.

[0140] The embodiments of the application also include a computer program product, which includes a computer program containing program codes for executing the method shown in the flow chart. When the computer program product is run in a computer system, the program codes are used to make the computer system implement the account processing method provided by the embodiments of the application.

[0141] The above functions defined in the system / apparatus of the embodiments of the application are performed when the computer program is executed by the processor 501. According to the embodiments of the application, the above-described system, apparatus, module, unit, etc. can be implemented by computer program modules.

[0142] In one embodiment, the computer program can rely on a tangible storage medium such as an optical storage device, a magnetic storage device, etc. In another embodiment, the computer program can also be transmitted, distributed, and downloaded in the form of a signal on a network medium, and be downloaded and installed through the communication part 509, and / or installed from the detachable medium 511. The program codes contained in the computer program can be transmitted by any appropriate network medium, including but not limited to wireless, wired, etc., or any appropriate combination thereof.

[0143] In such embodiments, the computer program can be downloaded and installed from the network through the communication section 509, and / or installed from the removable media 511. When the computer program is executed by the processor 501, the above-described functions defined in the system of the embodiments of the present application are executed. According to the embodiments of the present application, the system, device, apparatus, module, unit, and the like described above can be implemented by the computer program modules.

[0144] According to the embodiments of the present application, the program code for executing the computer program provided by the embodiments of the present application can be written in any combination of one or more programming languages, and specifically, these computer programs can be implemented using high-level procedural and / or object-oriented programming language, and / or assembly / machine language. The programming language includes, but is not limited to, such as Java, C++, python, "C" language, or similar programming language. The program code can be executed entirely on the user computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In the case involving a remote computing device, the remote computing device can be connected to the user computing device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, connected to the Internet through an Internet service provider).

[0145] The flowcharts and block diagrams in the drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowcharts or block diagrams can represent a module, a program segment, or a portion of code that contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in a different order than that shown in the figures. For example, two blocks noted in succession can actually be executed substantially concurrently or in the opposite order, depending on the functionality involved. It should also be noted that each block in the flowcharts or block diagrams, and combinations of blocks in the flowcharts or block diagrams, can be implemented by dedicated hardware-based systems that perform the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0146] Those skilled in the art can understand that the features described in each embodiment of the present application can be combined and / or integrated in various combinations, even if such combinations or integrations are not explicitly described in the present application. In particular, the features described in each embodiment of the present application can be combined and / or integrated in various combinations without departing from the spirit and teachings of the present application. All such combinations and / or integrations fall within the scope of the present application.

Claims

1. An account processing method, applied to a first processor, characterized in that, The method includes: Retrieve multiple risk accounts stored in the target database, determine the owner of each risk account, and obtain a set of owners; the risk accounts are those with abnormal transactions. Obtain transaction behavior data for each entity in the set of entities to which the entities belong; Based on the transaction behavior data, determine the risk threshold of the attribution subject set; The transaction behavior data of each attributable entity and the risk threshold are sent to the second processor; In response to receiving feedback information from the second processor, each risk account is processed according to the feedback information, wherein the feedback information is a processing strategy for each subject obtained by the second processor based on a pre-trained processing strategy model, which processes the transaction behavior data of each subject and the risk threshold.

2. The method according to claim 1, characterized in that, The step of obtaining transaction behavior data for each attributable entity in the attributable entity set includes: For any target owner in the set of owner entities, determine at least one account with the same name belonging to the target owner entity; Obtain transaction behavior data of at least one account with the same name, and use it as the transaction behavior data of the target entity.

3. The method according to claim 2, characterized in that, The acquisition of transaction behavior data of the at least one account with the same name includes: Obtain the duration of abnormal transactions for the risky account among the at least one account with the same name; Obtain the number of transactions of the at least one account with the same name within a preset time period; Obtain the transaction amount of the risky account among the at least one account with the same name.

4. The method according to claim 3, characterized in that, The step of determining the risk threshold of the attribution subject set based on the transaction behavior data includes: The median duration of the abnormal transaction, the number of transactions, and the transaction amount for the set of attributable entities are calculated respectively to obtain a first time threshold, a second number threshold, and a third amount threshold. The first time threshold, the second number threshold, and the third amount threshold are collectively determined as the risk threshold for the set of attributable entities.

5. The method according to claim 1, characterized in that, The feedback information includes a first processing strategy for the first attributing entity and a second processing strategy for the second attributing entity, wherein the number of risk accounts belonging to the first attributing entity is greater than the number of risk accounts belonging to the second attributing entity. The process of handling each risky account based on the feedback information includes: The risk account belonging to the first attributable entity is processed in accordance with the first processing strategy; the first processing strategy is used to instruct the risk account to be processed using a predetermined first execution action within the first execution channel; The risk accounts belonging to the second attributable entity are processed in accordance with the second processing strategy; the second processing strategy is used to instruct the risk accounts to be processed using a predetermined second execution action within the second execution channel; The first execution channel and the second execution channel are different, and the execution intensity corresponding to the first execution action is greater than the execution intensity corresponding to the second execution action.

6. The method according to claim 5, characterized in that, The method further includes: If no feedback information is received from the second processor within a preset time period, each risk account is processed according to the third processing strategy; The third processing strategy is used to instruct each risk account to be processed using a predetermined third execution action within the third execution channel. The execution intensity of the third execution action is greater than that of the second execution action and less than that of the first execution action.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: In response to receiving evaluation information on the processing results for each risk account, the evaluation information is sent to the second processor so that the second processor can retrain the pre-trained processing strategy model based on the evaluation information; The evaluation information includes positive evaluation information and negative evaluation information.

8. An account processing device, characterized in that, The device includes: The first determining module is used to obtain multiple risk accounts stored in the target database, determine the owner of each risk account, and obtain a set of owners; the risk account is an account with abnormal transactions. The acquisition module is used to acquire transaction behavior data for each of the homeowners in the homeowner set; The second determining module is used to determine the risk threshold of the attribution subject set based on the transaction behavior data; The sending module is used to send the transaction behavior data of each attributing entity and the risk threshold to the second processor; The processing module is configured to respond to receiving feedback information from the second processor and process each risk account according to the feedback information, wherein the feedback information is a processing strategy for each subject obtained by the second processor based on a pre-trained processing strategy model, which processes the transaction behavior data of each subject and the risk threshold.

9. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that 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, characterized in that, 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, characterized in that, 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