Method and device for automatically recharging payment account, electronic equipment and medium

Abstract

The invention provides a method and a device for automatically recharging a payment account, electronic equipment and a medium, which can be applied to the technical field of big data and the technical field of artificial intelligence. The method comprises the following steps: acquiring a payment account; based on the payment account, binding a bank account corresponding to the payment account; based on a federated prediction model, a predicted payment amount and predicted payment time of the payment account are obtained, and the federated prediction model is obtained based on a federated learning framework and by using historical payment information of the payment account as training data through pre-training; judging an automatic recharging opportunity based on the predicted payment time; in response to the arrival of the automatic recharging opportunity, selecting a pre-payment account from bank accounts bound with the payment account; and transferring funds from the pre-payment account to the payment account based on the predicted payment amount so as to carry out automatic recharging.

Description

Technical Field

[0001] The present invention relates to the fields of big data technology and artificial intelligence technology, and more specifically to a method, device, electronic device and medium for automatically recharging a payment account. Background Art

[0002] With the continuous advancement of technology and changes in lifestyle, life payment services are also constantly evolving and improving. In the traditional payment process, users usually need to estimate the amount to be paid in advance before the payment bill arrives and manually pay and recharge.

[0003] However, this method may have some inconveniences and limitations in actual operation, and may lead to the following problems: First, users need to estimate future payment amounts, but consumption in life is often changeable and difficult to predict accurately, which may lead to excessive or insufficient recharge amounts, affecting the user's fund utilization efficiency; second, users need to actively remember the payment date. If they forget or miss the payment, it may result in overdue fees or service interruption; third, the traditional model cannot make intelligent adjustments based on individual users' consumption habits and expenditures, and lacks personalized services.

[0004] Therefore, it is necessary to establish a smarter and more adaptive payment method to provide users with a more convenient, accurate and personalized payment experience. Summary of the invention

[0005] In view of the above problems, according to a first aspect of the present invention, a method for automatic recharge of a payment account is provided, the method comprising: obtaining a payment account; based on the payment account, binding a bank account corresponding to the payment account; based on a federal prediction model, obtaining a predicted payment amount and a predicted payment time of the payment account, wherein the federal prediction model is based on a federated learning framework and is pre-trained using the historical payment information of the payment account as training data; based on the predicted payment time, determining an automatic recharge timing; in response to reaching the automatic recharge timing, selecting a prepaid account from the bank account bound to the payment account; and based on the predicted payment amount, transferring funds from the prepaid account to the payment account for automatic recharge.

[0006] According to some exemplary embodiments, the pre-trained federated prediction model specifically includes: establishing a federated learning framework, the federated learning framework includes a server side and n clients, wherein the n clients include a bank side and a charging enterprise side, the bank side and the charging enterprise side include respective historical payment information databases, wherein n is a positive integer greater than 1; initializing basic models at the n clients respectively, and using the data in the respective historical payment information databases of the n clients as training data to perform local model training respectively, and obtain the first model parameters of the n clients; parameter aggregation operation: the n clients upload their respective first model parameters to the server side, and the server side uploads the n clients to the server side. The first model parameters of the client are aggregated to obtain the second model parameters; model update operation: the server sends the second model parameters to the n clients, and the n clients re-train the local model based on the second model parameters and the historical payment information database, and update the first model parameters of the n clients; parameter re-aggregation operation: the n clients upload the updated first model parameters to the server, and the server aggregates the updated first model parameters of the n clients to obtain the updated second model parameters; and repeat the parameter aggregation operation, the model update operation and the parameter re-aggregation operation until the model converges or the maximum number of iterations is reached.

[0007] According to some exemplary embodiments, the server aggregates the first model parameters of the n clients in a weighted average manner, wherein the weights of the first model parameters of the n clients are determined according to the data amounts of the n clients.

[0008] According to some exemplary embodiments, in the model updating operation, the first model parameters are updated using empirical risk minimization.

[0009] According to some exemplary embodiments, binding a bank account corresponding to the payment account based on the payment account specifically includes: obtaining identity authentication information of the payment account; matching the identity authentication information with bank account information in multiple bank account databases using an information matching method; and in response to a successful match, binding the payment account with the bank account.

[0010] According to some exemplary embodiments, the information matching method is used to match the identity authentication information with bank account information in multiple bank account databases, specifically including: preprocessing the identity authentication information; selecting matching criteria based on the data format of the bank account database; and selecting a target matching algorithm to search for matching accounts in the bank account database, wherein the target matching algorithm includes: string matching, edit distance algorithm or hash function.

[0011] According to some exemplary embodiments, the selecting a prepaid account from the bank accounts corresponding to the payment account specifically includes: traversing the bank accounts bound to the payment account, comparing the balances, and selecting the bank account with the largest balance; or traversing the bank accounts bound to the payment account, calculating the difference between the balance of each bank account and the predicted payment amount, and selecting the bank account with the smallest difference; or selecting the bank account with the highest selection weight according to the pre-set selection weights of each bank account bound to the payment account.

[0012] According to a second aspect of the present invention, a device for automatically recharging a payment account is proposed, the device comprising: a payment account acquisition module, used to: acquire a payment account; a binding module, used to: bind a bank account corresponding to the payment account based on the payment account; a prediction data acquisition module, used to: acquire a predicted payment amount and a predicted payment time of the payment account based on a federal prediction model, wherein the federal prediction model is based on a federated learning framework and is pre-trained using the historical payment information of the payment account as training data; a judgment module, used to: judge an automatic recharge timing based on the predicted payment time; a prepaid account selection module, used to: select a prepaid account from the bank account corresponding to the payment account in response to reaching the automatic recharge timing; and an automatic recharge module, used to: transfer funds from the prepaid account to the payment account based on the predicted payment amount for automatic recharge.

[0013] According to some exemplary embodiments, the binding module may include a bank account binding module.

[0014] According to some exemplary embodiments, the bank account binding module may include an identity authentication information acquisition unit, an information matching module and a binding unit.

[0015] According to some exemplary embodiments, the identity authentication information acquisition unit may be used to acquire the identity authentication information of the payment account.

[0016] According to some exemplary embodiments, the information matching module may be used to match the identity verification information with bank account information in a plurality of bank account databases using an information matching method.

[0017] According to some exemplary embodiments, the binding unit may be configured to bind the payment account with a bank account in response to a successful match.

[0018] According to some exemplary embodiments, the information matching module may include a pre-processing unit, a matching criterion selection unit, and a target matching algorithm selection unit.

[0019] According to some exemplary embodiments, the preprocessing unit may be configured to perform data preprocessing on the identity authentication information.

[0020] According to some exemplary embodiments, the matching criterion selection unit may be configured to select a matching criterion based on a data format of a bank account database.

[0021] According to some exemplary embodiments, the target matching algorithm selection unit may be used to select a target matching algorithm to search for matching accounts in a bank account database, wherein the target matching algorithm includes: string matching, edit distance algorithm or hash function.

[0022] According to some exemplary embodiments, the prediction data acquisition module may include a federated prediction model training module.

[0023] According to some exemplary embodiments, the federated prediction model training module may include a federated learning framework building unit, a basic model initialization unit, a parameter aggregation unit, a model updating unit, a parameter re-aggregation unit, and an iteration unit.

[0024] According to some exemplary embodiments, the federated learning framework establishing unit can be used to establish a federated learning framework, which includes a server side and n clients, wherein the n clients include a bank side and a charging enterprise side, and the bank side and the charging enterprise side include respective historical payment information databases, wherein n is a positive integer greater than 1.

[0025] According to some exemplary embodiments, the basic model initialization unit can be used to initialize the basic models at the n clients respectively, and use the data in the historical payment information database of each of the n clients as training data to perform local model training respectively, so as to obtain the first model parameters of the n clients.

[0026] According to some exemplary embodiments, the parameter aggregation unit may be used for parameter aggregation operation: the n clients upload their respective first model parameters to the server, and the server aggregates the first model parameters of the n clients to obtain second model parameters.

[0027] According to some exemplary embodiments, the model update unit can be used for model update operations: the server sends the second model parameters to the n clients, and the n clients re-train local models based on the second model parameters and the historical payment information database, respectively, to update the first model parameters of the n clients.

[0028] According to some exemplary embodiments, the parameter reaggregation unit can be used for parameter reaggregation operation: the n clients respectively upload the updated first model parameters to the server, and the server aggregates the updated first model parameters of the n clients to obtain updated second model parameters.

[0029] According to some exemplary embodiments, the iteration unit may be configured to repeat the parameter aggregation operation, the model updating operation, and the parameter re-aggregation operation until the model converges or a maximum number of iterations is reached.

[0030] According to some exemplary embodiments, the prepayment account selection module may include a prepayment account method selection module.

[0031] According to some exemplary embodiments, the prepayment account method selection module may include a maximum balance selection unit, a minimum difference selection unit, or a highest weight selection unit.

[0032] According to some exemplary embodiments, the maximum balance selection unit may be configured to traverse the bank accounts bound to the payment account, compare the balances, and select the bank account with the largest balance.

[0033] According to some exemplary embodiments, the minimum difference selection unit may be used to traverse the bank accounts bound to the payment account, calculate the difference between the balance of each bank account and the predicted payment amount, and select the bank account with the smallest difference.

[0034] According to some exemplary embodiments, the highest weight selection unit may be configured to select the bank account with the highest selection weight according to a pre-set selection weight for each bank account bound to the payment account.

[0035] According to a third aspect of the present invention, there is provided an electronic device, comprising: one or more processors; a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors execute the method as described above.

[0036] According to a fourth aspect of the present invention, there is provided a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, causes the processor to execute the method as described above.

[0037] According to a fifth aspect of the present invention, there is provided a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method as described above is implemented.

[0038] One or more of the above-mentioned embodiments have the following advantages or beneficial effects: the method for automatic recharging of a payment account provided by the present invention can predict future payment needs and amounts based on the user's historical behavior and consumption patterns, realize an intelligent automatic recharging process, reduce the user's operation and waiting time, and improve the user experience; at the same time, by adopting federal computing technology, data processing and analysis are distributed on different data platforms, and the original data is not directly exchanged, thereby protecting the user's privacy and sensitive information. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above contents and other objects, features and advantages of the present invention will become more apparent through the following description of the embodiments of the present invention with reference to the accompanying drawings, in which:

[0040] Figure 1 The application scenario diagram of the method, apparatus, device, and medium for automatically recharging a payment account according to an embodiment of the present invention is schematically shown.

[0041] Figure 2 The flowchart of the method for automatic recharging of a payment account according to an embodiment of the present invention is schematically shown.

[0042] Figure 3 The flowchart of the method for binding a bank account according to an embodiment of the present invention is schematically shown.

[0043] Figure 4 The flowchart of the method for binding a bank account according to an embodiment of the present invention is schematically shown.

[0044] Figure 5 A flowchart of a method for pre-training a federated prediction model according to an embodiment of the present invention is schematically shown.

[0045] Figure 6 The flowchart of the method for selecting a prepayment account according to an embodiment of the present invention is schematically shown.

[0046] Figure 7 The structure block diagram of the apparatus for automatically recharging a payment account according to an embodiment of the present invention is schematically shown.

[0047] Figure 8 A block diagram of an electronic device suitable for the method for automatic recharging of a payment account according to an embodiment of the present invention is schematically shown. DETAILED DESCRIPTION

[0048] Below, 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 present invention. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of embodiments of the present invention. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of concepts of the present invention.

[0049] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0050] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0051] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0052] In the technical solution of the present invention, the acquisition, storage and application of user personal information involved all comply with the provisions of relevant laws and regulations, take necessary confidentiality measures, and do not violate public order and good morals.

[0053] First, the technical terms described in this article are explained and described as follows.

[0054] Federated Learning is a distributed machine learning method that aims to train machine learning models while protecting the data privacy of participants. Unlike traditional centralized machine learning, the core idea of ​​federated learning is to move the model training process to where the data is, avoiding storing the data in one place, thereby achieving collaborative training of the model without sharing the original data. In federated learning, there is a central server and multiple clients. Each client has its own local data, which may be different users' devices or databases of different institutions in a distributed environment.

[0055] With the popularity of online third-party payment tools, people no longer need to go to the business hall of offline charging companies to handle payment services. They can use various payment tools to pay the payment bills of charging companies. In traditional payment services, users need to estimate the amount to be paid in advance before the payment bill date arrives, and manually complete the payment recharge. This process not only involves the complexity of user prediction and payment, but also requires manual operation within a specific time, which may lead to delays and cumbersome payment experience. However, with the continuous development and popularization of technology, all walks of life are actively exploring how to use information technology to optimize traditional business processes, improve user experience and reduce the complexity of operations.

[0056] In the current digital age, innovations such as mobile payments, electronic bills, and automated services are leading the transformation of the payment and financial sectors. For example, the system can provide real-time notifications, automatic accounting, bill analysis, and other functions to improve user satisfaction and payment efficiency. However, for life payment services, such as utility bills, communication fees, etc., there are some potential problems and inconveniences in the user's need to estimate the payment amount and the manual payment process. First, users need to remember the payment time, estimate the amount, and pay manually, which increases the user's burden. Especially when multiple bills need to be processed, the operation is cumbersome. Second, manually completing the payment requires the user's time and attention, which may be overlooked or omitted in busy daily life. Third, the user's estimate of the future payment amount may be inaccurate, resulting in insufficient or excessive recharge. This may cause bills to be not paid on time or funds to be idle in the account, reducing the liquidity of funds.

[0057] Based on this, an embodiment of the present invention provides a method for automatically recharging a payment account, comprising: obtaining a payment account; binding a bank account corresponding to the payment account based on the payment account; obtaining a predicted payment amount and a predicted payment time of the payment account based on a federated prediction model, wherein the federated prediction model is pre-trained based on a federated learning framework and using the historical payment information of the payment account as training data; judging the timing of automatic recharging based on the predicted payment time; selecting a prepaid account from the bank account bound to the payment account in response to reaching the automatic recharge timing; and transferring funds from the prepaid account to the payment account based on the predicted payment amount for automatic recharging. The method for automatically recharging a payment account provided by the present invention can predict future payment needs and amounts based on user historical behaviors and consumption patterns, realize intelligent automatic recharging processes, reduce user operations and waiting time, and improve user experience; at the same time, by using federated computing technology, data processing and analysis are distributed on different data platforms, and original data is not directly exchanged, thereby protecting user privacy and sensitive information.

[0058] It should be noted that the method, device, equipment and medium for automatic recharging of a payment account determined by the present invention can be used in the fields of big data technology and artificial intelligence technology, can also be used in the financial field, and can also be used in various fields other than the fields of big data technology and artificial intelligence technology and the financial field. The application fields of the method, device, equipment and medium for automatic recharging of a payment account provided by the embodiments of the present invention are not limited.

[0059] In the technical solution of the present invention, 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 used for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data comply with the relevant laws, regulations and standards of relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0060] Figure 1 The application scenario diagram of the method, apparatus, device, and medium for automatically recharging a payment account according to an embodiment of the present invention is schematically shown.

[0061] like Figure 1 As shown, the application scenario 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for a communication link between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0062] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only examples).

[0063] The terminal devices 101 , 102 , and 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, and desktop computers.

[0064] The server 105 may be a server that provides various services, such as a background management server (only an example) that provides support for websites browsed by users using the terminal devices 101, 102, and 103. The background management server may analyze and process the received data such as user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.

[0065] It should be noted that the method for automatically recharging a payment account provided in the embodiment of the present invention can generally be executed by the server 105. Accordingly, the device for automatically recharging a payment account provided in the embodiment of the present invention can generally be set in the server 105. The method for automatically recharging a payment account provided in the embodiment of the present invention can also be executed by a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and / or the server 105. Accordingly, the device for automatically recharging a payment account provided in the embodiment of the present invention can also be set in a server or server cluster that is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and / or the server 105.

[0066] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0067] Figure 2 The flowchart of the method for automatic recharging of a payment account according to an embodiment of the present invention is schematically shown.

[0068] like Figure 2 As shown, the method 200 for automatically recharging a payment account in this embodiment may include operations S210 to S260.

[0069] In operation S210, a payment account is obtained.

[0070] In an embodiment of the present invention, a user can authenticate his / her identity and register a payment account on a three-party life payment platform built by a bank for charging enterprises and payers. After successful registration and verification, a unique customer number will be obtained and a payment account corresponding to the customer number will be generated.

[0071] In the embodiment of the present invention, the payment account may include multiple types according to different service types and requirements, such as water bill account, electricity bill account, phone bill account, property management fee account, etc. The third-party life payment platform usually provides bill query and management functions for users to view historical bills, payment records and other information.

[0072] In operation S220, based on the payment account, a bank account corresponding to the payment account is bound.

[0073] In an embodiment of the present invention, the system can adaptively match and bind the bank account with the payment account. In this way, the system can automatically transfer funds from the bank account to the payment account according to the needs of the user, thereby realizing automatic recharge.

[0074] Figure 3 The flowchart of the method for binding a bank account according to an embodiment of the present invention is schematically shown.

[0075] like Figure 3 As shown, the method for binding a bank account in this embodiment may include operations S310 to S330, and the operations S310 to S330 may at least partially execute the above-mentioned operation S220.

[0076] In operation S310, identity authentication information of the payment account is obtained.

[0077] In an embodiment of the present invention, the user needs to complete the registration on the platform through identity verification information, which will provide a basis for the subsequent adaptive matching of bank accounts. The identity verification information of the payment account usually includes the user's identity, password, or other verification credentials, which are key factors for confirming the user's identity and authorizing access.

[0078] Among them, adaptive matching aims to dynamically select or adaptively match the most appropriate solutions, options or strategies according to different situations and needs. It emphasizes flexible matching according to current conditions and contexts to obtain the best results. Adaptive matching has applications in various fields, including data processing, decision making, resource allocation, etc. Specifically, adaptive matching can adjust matching strategies or select solutions in real time according to changes in the environment and actual needs. It can flexibly respond to different situations to achieve the best or most appropriate effect. At the same time, adaptive matching can also select the most suitable options for their needs according to individual characteristics and preferences. This helps to provide personalized experiences and solutions. In complex situations, which often involve multiple variables and factors, adaptive matching can make intelligent choices based on these variables, thereby achieving good results in complex environments. Adaptive matching is real-time and can quickly adapt to new situations and changes, so as to make appropriate choices quickly.

[0079] In operation S320, the identity verification information is matched with bank account information in a plurality of bank account databases using an information matching method.

[0080] In an embodiment of the present invention, once the payment account is successfully registered, the system will use the identity verification information of the payment account to automatically try to match information in multiple bank account databases. The system will search for matching bank account information in the account databases of different banks based on the identity verification information provided by the payment account, which may involve comparing information such as name, ID number, mobile phone number, etc. to ensure the accuracy of the match.

[0081] Figure 4 The flowchart of the method for binding a bank account according to an embodiment of the present invention is schematically shown.

[0082] like Figure 4 As shown, the method for binding a bank account in this embodiment may include operations S410 to S430, and the operations S410 to S430 may at least partially execute the above-mentioned operation S320.

[0083] In operation S410, the identity verification information is subjected to data preprocessing.

[0084] In an embodiment of the present invention, in order to provide clean and standardized data in the subsequent matching process, improve the accuracy and reliability of matching, and reduce errors caused by data format differences, specific preprocessing operations may include the following aspects: text case unification. The text in the input data may be mixed case. In order to avoid differences caused by case during the matching process, the text can be uniformly converted to all uppercase or all lowercase; remove special characters. The input data may contain various special characters (such as punctuation marks, blank characters, etc.), which may interfere with the matching process. In the preprocessing process, special characters can be considered to be removed or replaced; format conversion. If the input data involves information in a specific format such as numbers and dates, it can be converted into a unified format to ensure that the data is comparable during matching; and remove redundant information. The input data may contain some redundant information that is not related to matching. The preprocessing process can remove this information to reduce interference.

[0085] In operation S420, a matching criterion is selected based on the data format of the bank account database.

[0086] In an embodiment of the present invention, different bank account databases have different data formats. Therefore, it is necessary to determine the criteria or fields for matching according to the data format stored in the bank account database so as to make a correct comparison during the matching process. Specifically, the following processes may be included: understanding of the data format. The data in the bank account database may contain different fields, such as account name, account number, mobile phone number, email address, etc. Before matching, it is necessary to understand the data format stored in the database and understand the meaning and format of each field; based on the understanding of the data format, select the criteria for matching. This can be a specific field or a combination of multiple fields. For example, if a bank account database uses account name and mobile phone number as identifiers, and the identity verification information provides the account name and mobile phone number, then these two fields can be selected as matching criteria.

[0087] In an embodiment of the present invention, when selecting matching criteria, it is necessary to ensure that the selected fields have sufficient uniqueness and accuracy. If the selected fields are not unique enough, the matching results may be uncertain or inaccurate. For example, when selecting an account name as a criterion, it is necessary to ensure that the account name is unique in the database. Furthermore, when selecting matching criteria, it is also necessary to pay attention to whether the format of the fields is uniform. If the format of the data in different records is inconsistent, data cleaning or formatting may be required to ensure the accuracy of the match. In addition, when selecting matching criteria, fault tolerance should also be considered, that is, even if there are slight differences in the data, the match can still be performed correctly, which can be achieved by using methods such as fuzzy matching.

[0088] In operation S430, a target matching algorithm is selected to search for matching accounts in a bank account database, wherein the target matching algorithm includes: string matching, edit distance algorithm or hash function.

[0089] In the embodiment of the present invention, after selecting a matching standard, a corresponding matching algorithm needs to be adapted according to the selected standard, and different standards may require different matching methods for comparison.

[0090] In the embodiment of the present invention, string matching is a basic matching algorithm that matches by accurately comparing the content of the string; if the identity authentication information and the account information in the bank account database have the same format and do not contain errors, string matching can be selected as an effective method; the edit distance algorithm can measure the similarity between two strings, that is, the minimum steps required to convert one string into another string through insertion, deletion and substitution operations, which may be more fault-tolerant when processing data with slight differences; the hash function can map the input data to a hash value of a fixed length. If there is account information processed using a hash function in the bank account database, matching can be achieved by applying the same hash function to the identity authentication information and searching for matching hash values ​​in the database.

[0091] Return to reference Figure 3 In operation S330, in response to a successful match, the payment account is bound to the bank account.

[0092] In an embodiment of the present invention, if the information matches successfully, that is, the system finds bank account information that matches the bill payment account identity authentication information, the system will respond successfully and bind the bill payment account to the bank account. This means that the user can connect his bill payment account to the bank account so that funds can be transferred from the bank account to the bill payment account for paying bills.

[0093] It should be noted that the identity authentication information of the payment account is sensitive information, so it is necessary to ensure the security of the communication process and the privacy protection of stored and processed data; at the same time, when processing user data and bank account information, it is necessary to comply with relevant laws and regulations, such as data protection laws and financial privacy regulations.

[0094] It should be noted that for matching failures or other problems, the system can provide appropriate error handling mechanisms so that users can understand the cause of the problem and take corrective measures.

[0095] Return to reference Figure 2 In operation S230, based on the federated prediction model, the predicted payment amount and predicted payment time of the payment account are obtained, wherein the federated prediction model is pre-trained based on a federated learning framework and using the historical payment information of the payment account as training data.

[0096] Figure 5 A flowchart of a method for pre-training a federated prediction model according to an embodiment of the present invention is schematically shown.

[0097] like Figure 5As shown, the method for pre-training a federated prediction model in this embodiment may include operations S510 to S560, and operations S510 to S560 may at least partially execute the above-mentioned operation S230.

[0098] In operation S510, a federated learning framework is established, which includes a server side and n clients, wherein the n clients include a bank side and a charging enterprise side, and the bank side and the charging enterprise side include respective historical payment information databases, wherein n is a positive integer greater than 1.

[0099] In an embodiment of the present invention, the federated learning framework includes a server and n clients, and the server is responsible for coordinating and managing the entire federated learning process. The historical payment information database of each of the n clients includes the historical payment data of the payment account obtained through data collection, including but not limited to the payment amount, payment time, number of payments, payment items, etc.

[0100] In operation S520, basic models are initialized at the n clients respectively, and local model training is performed respectively using data in the historical payment information database of each of the n clients as training data to obtain first model parameters of the n clients.

[0101] In an embodiment of the present invention, each client (e.g., a bank account or a charging enterprise) initializes a basic model locally, and selecting an appropriate basic model architecture and parameters is crucial to the success of subsequent training. Considering that each client may have different data characteristics, it may be necessary to select a flexible model, such as a multi-layer architecture of a neural network, so that the model can capture patterns in different data distributions.

[0102] In an embodiment of the present invention, each client uses the local historical payment information database as training data to independently train its basic model. Therefore, the bank account end and the charging enterprise end use their own data to train the model, which can ensure that data privacy is not leaked.

[0103] In operation S530, parameter aggregation operation: the n clients upload their respective first model parameters to the server, and the server aggregates the first model parameters of the n clients to obtain second model parameters.

[0104] In an embodiment of the present invention, the server may aggregate the first model parameters of the n clients in a weighted average manner, wherein the weights of the first model parameters of the n clients are determined according to the data amounts of the n clients.

[0105] In an embodiment of the present invention, weighted average is a commonly used parameter aggregation method, and its basic idea is to assign different weights according to the amount of data of each client, so as to better reflect the contribution of each client in the global model. Specifically, the server side can calculate the corresponding weight according to the amount of data of each client. Clients with larger data volumes will have greater weights in the aggregation because they can provide more information. This weight calculation can be achieved through a simple ratio of the number of data samples, or other more complex weight allocation strategies can be adopted, such as considering differences in client data distribution.

[0106] In an embodiment of the present invention, weighted averaging can better handle the situation of data imbalance, ensuring that clients with larger data volumes contribute more to the global model, thereby avoiding model deviations caused by too little data from individual clients.

[0107] In operation S540, the model update operation: the server sends the second model parameters to the n clients, and the n clients re-train local models based on the second model parameters and the historical payment information database, respectively, to update the first model parameters of the n clients.

[0108] In an embodiment of the present invention, the server can send the second model parameter back to each client. Each client uses this parameter and its local historical payment information database to update the local model. This will help the client to fine-tune the model locally to adapt to a specific data distribution.

[0109] It should be noted that the number of training iterations and learning rate can be controlled during the model update operation to avoid excessive fitting of local data. At the same time, the client can use variants of the optimization algorithm to adapt to the characteristics of distributed training, such as using distributed SGD (stochastic gradient descent).

[0110] In an embodiment of the present invention, in the model update operation, the first model parameters are updated using empirical risk minimization, which is a commonly used machine learning optimization strategy that aims to minimize the loss function of the model on local data, thereby making the model better fit the local data distribution.

[0111] In the embodiment of the present invention, it is assumed that in the tth iteration, the local model parameters of the ith participant are \boldsymbol{w}_i^{(t)}. During local update, the ith participant will update its local model parameters according to its local data set Di, which can be achieved by minimizing the empirical risk:

[0112]

[0113] In operation S550, parameter re-aggregation operation: the n clients upload the updated first model parameters to the server respectively, and the server aggregates the updated first model parameters of the n clients to obtain updated second model parameters.

[0114] In an embodiment of the present invention, the client can upload the updated first model parameters to the server. The server aggregates the parameters from all clients again to obtain updated global model parameters (second model parameters).

[0115] In an embodiment of the present invention, the participants need to upload their locally updated model parameters to the server for global aggregation in order to obtain more accurate model parameters. Assume that after the tth iteration, the local model parameters of all participants are {\boldsymbol{w}_1^{(t+1)}, \boldsymbol{w}_2^{(t+1)}, …, \boldsymbol{w}_n^{(t+1)}}. Then global aggregation can be achieved by the following formula:

[0116]

[0117] Wherein, \boldsymbol{w}^{(t+1)} is the global model parameter after the t+1th iteration, |\mathcal{D}i| represents the size of the dataset owned by participant i, and |\mathcal{D}|=\sum{i=1}^n|\mathcal{D}_i| represents the sum of the dataset sizes of all participants.

[0118] In operation S560, the parameter aggregation operation, the model updating operation, and the parameter re-aggregation operation are repeated until the model converges or a maximum number of iterations is reached.

[0119] It should be noted that the key advantage of the federated learning framework is to protect data privacy. Since the original data does not leave the client, only the model parameters are aggregated on the server side, and the user's data is protected. In addition, the model is trained on the local data of different clients, which can adapt to diverse data distributions and improve the generalization performance of the model.

[0120] Return to reference Figure 2 In operation S240, based on the predicted payment time, the timing of automatic recharge is determined.

[0121] In an embodiment of the present invention, based on the predicted payment time, the system can determine when is the best time for automatic recharge based on the user's historical records. In addition, the conditions for triggering automatic recharge can also be set based on user needs, such as when the predicted payment time is less than a certain number of days.

[0122] In operation S250, in response to reaching the automatic recharge opportunity, a prepaid account is selected from the bank account bound to the payment account.

[0123] In an embodiment of the present invention, a method for adaptively selecting a prepayment account from a plurality of bank accounts bound to a payment account may be implemented based on user preferences, historical data, environment, and other factors.

[0124] Figure 6 The flowchart of the method for selecting a prepayment account according to an embodiment of the present invention is schematically shown.

[0125] like Figure 6 As shown, the method for selecting a prepayment account in this embodiment may include operation S610 or S620 or S630, and the operation S610 or S620 or S630 may at least partially execute the above operation S250.

[0126] In operation S610, the bank accounts bound to the payment account are traversed, the balances are compared, and the bank account with the largest balance is selected.

[0127] In an embodiment of the present invention, the bank account with the largest balance can be selected as the prepayment account by comparing the balances of the bank accounts bound to the payment account. This method is suitable for the case where the user wants to pay from the most sufficient bank account.

[0128] In operation S620, the bank accounts bound to the payment account are traversed, the difference between the balance of each bank account and the predicted payment amount is calculated, and the bank account with the smallest difference is selected.

[0129] In an embodiment of the present invention, the difference between the balance of each bank account and the predicted payment amount can be calculated, and the bank account with the smallest difference can be selected as the prepayment account. This takes into account the predicted payment amount to reduce the gap between the balance and the expected expenditure.

[0130] In operation S630, according to the preset selection weights of the bank accounts bound to each of the payment accounts, the bank account with the highest selection weight is selected.

[0131] In an embodiment of the present invention, the bank account with the highest selection weight can be selected as the prepayment account according to the pre-set selection weight of the bank account bound to each payment account. This method allows users to adjust the selection of prepayment accounts according to their preferences and needs.

[0132] Return to reference Figure 2 , in operation S260, based on the predicted payment amount, funds are transferred from the prepaid account to the payment account for automatic recharge.

[0133] In an embodiment of the present invention, after the funds transfer is completed, the system can update the balance information of the prepaid account and the bill payment account to reflect the change in the funds transfer. Over time, the system can optimize the prediction model, recharge threshold and other strategies based on user feedback and recharge conditions to provide a more accurate recharge experience that meets user needs.

[0134] The method for automatic recharging of payment accounts provided by the present invention can predict future payment needs and amounts based on the user's historical behavior and consumption patterns, realize intelligent automatic recharging process, reduce user operation and waiting time, and improve user experience; at the same time, the federated computing technology is used to distribute data processing and analysis on different data platforms, without directly exchanging original data, thus protecting the user's privacy and sensitive information. Specifically, it brings the following benefits:

[0135] Beneficial effects:

[0136] 1. Federated computing technology is used to distribute data processing and analysis on different data platforms, without directly exchanging original data, thus protecting the privacy and sensitive information of users. All parties jointly conduct model training and analysis while maintaining data privacy, thus improving data security;

[0137] 2. Through automated and intelligent processes, the recharge process of the payment account is simplified into a series of automatic operations. Users do not need to frequently log in, check balances or recharge manually, thus improving the convenience of payment;

[0138] 3. It can predict future payment needs and amounts based on users’ historical behaviors and consumption patterns, realize intelligent automatic recharge process, reduce users’ operation and waiting time, and improve user experience;

[0139] 4. The automated recharge process reduces the possibility of user errors, such as entering the wrong amount or account information, and, if a weight-based approach is adopted, users can personalize the settings according to their needs and preferences. This increases the user's control over the recharge process.

[0140] Based on the above-mentioned method for automatically recharging a payment account, the present invention also provides a device for automatically recharging a payment account. Figure 7 The device is described in detail.

[0141] Figure 7 The structure block diagram of the apparatus for automatically recharging a payment account according to an embodiment of the present invention is schematically shown.

[0142] like Figure 7As shown, the device 700 for automatically recharging a payment account according to this embodiment includes a payment account acquisition module 710, a binding module 720, a prediction data acquisition module 730, a judgment module 740, a prepayment account selection module 750 and an automatic recharging module 760.

[0143] The payment account acquisition module 710 may be used to acquire a payment account. In one embodiment, the payment account acquisition module 710 may be used to perform the operation S210 described above, which will not be described in detail herein.

[0144] The binding module 720 may be used to bind a bank account corresponding to the payment account based on the payment account. In one embodiment, the binding module 720 may be used to perform the operation S220 described above, which will not be described in detail herein.

[0145] The predicted data acquisition module 730 can be used to obtain the predicted payment amount and predicted payment time of the payment account based on the federated prediction model, wherein the federated prediction model is pre-trained based on the federated learning framework and using the historical payment information of the payment account as training data. In one embodiment, the predicted data acquisition module 730 can be used to perform the operation S230 described above, which will not be repeated here.

[0146] The determination module 740 may be used to determine the timing of automatic recharge based on the predicted payment time. In one embodiment, the determination module 740 may be used to perform the operation S240 described above, which will not be described in detail herein.

[0147] The prepayment account selection module 750 may be used to select a prepayment account from the bank account corresponding to the payment account in response to reaching the automatic recharge opportunity. In one embodiment, the prepayment account selection module 750 may be used to perform the operation S250 described above, which will not be described in detail here.

[0148] The automatic recharge module 760 may be used to transfer funds from the prepaid account to the payment account based on the predicted payment amount for automatic recharge. In one embodiment, the automatic recharge module 760 may be used to perform the operation S260 described above, which will not be described in detail here.

[0149] According to an embodiment of the present invention, the binding module 720 may include a bank account binding module.

[0150] According to an embodiment of the present invention, the bank account binding module may include an identity authentication information acquisition unit, an information matching module and a binding unit.

[0151] The identity verification information acquisition unit may be used to obtain the identity verification information of the payment account. In one embodiment, the identity verification information acquisition unit may be used to perform the operation S310 described above, which will not be described in detail here.

[0152] The information matching module may be used to match the identity authentication information with bank account information in multiple bank account databases using an information matching method. In one embodiment, the information matching module may be used to perform the operation S320 described above, which will not be described in detail herein.

[0153] The binding unit may be used to bind the payment account with the bank account in response to a successful match. In one embodiment, the binding unit may be used to perform the operation S330 described above, which will not be described in detail herein.

[0154] According to an embodiment of the present invention, the information matching module may include a pre-processing unit, a matching standard selection unit and a target matching algorithm selection unit.

[0155] The preprocessing unit may be used to perform data preprocessing on the identity verification information. In one embodiment, the preprocessing unit may be used to perform the operation S410 described above, which will not be described in detail herein.

[0156] The matching standard selection unit may be used to select a matching standard based on the data format of the bank account database. In one embodiment, the matching standard selection unit may be used to perform the operation S420 described above, which will not be described in detail herein.

[0157] The target matching algorithm selection unit can be used to select a target matching algorithm to search for matching accounts in a bank account database, wherein the target matching algorithm includes: string matching, edit distance algorithm or hash function. In one embodiment, the target matching algorithm selection unit can be used to perform the operation S430 described above, which will not be repeated here.

[0158] According to an embodiment of the present invention, the prediction data acquisition module 730 may include a federated prediction model training module.

[0159] According to an embodiment of the present invention, the federated prediction model training module may include a federated learning framework establishment unit, a basic model initialization unit, a parameter aggregation unit, a model updating unit, a parameter re-aggregation unit and an iteration unit.

[0160] The federated learning framework establishment unit may be used to establish a federated learning framework, the federated learning framework comprising a server end and n clients, wherein the n clients comprise a bank end and a charging enterprise end, the bank end and the charging enterprise end comprise respective historical payment information databases, wherein n is a positive integer greater than 1. In one embodiment, the federated learning framework establishment unit may be used to perform the operation S510 described above, which will not be described in detail herein.

[0161] The initialization basic model unit can be used to initialize the basic model at the n clients respectively, and use the data in the historical payment information database of each of the n clients as training data to perform local model training respectively, and obtain the first model parameters of the n clients. In one embodiment, the initialization basic model unit can be used to perform the operation S520 described above, which will not be repeated here.

[0162] The parameter aggregation unit can be used for parameter aggregation operation: the n clients upload their respective first model parameters to the server, and the server aggregates the first model parameters of the n clients to obtain second model parameters. In one embodiment, the parameter aggregation unit can be used to perform the operation S530 described above, which will not be repeated here.

[0163] The model updating unit can be used for a model updating operation: the server sends the second model parameters to the n clients, and the n clients respectively retrain local models based on the second model parameters and the historical payment information database to update the first model parameters of the n clients. In one embodiment, the model updating unit can be used to perform the operation S540 described above, which will not be described in detail here.

[0164] The parameter re-aggregation unit may be used for parameter re-aggregation operation: the n clients upload the updated first model parameters to the server respectively, and the server aggregates the first model parameters updated by the n clients to obtain updated second model parameters. In one embodiment, the parameter re-aggregation unit may be used to perform the operation S550 described above, which will not be described in detail here.

[0165] The iteration unit may be used to repeat the parameter aggregation operation, the model update operation and the parameter re-aggregation operation until the model converges or reaches a maximum number of iterations. In one embodiment, the iteration unit may be used to perform the operation S560 described above, which will not be described in detail herein.

[0166] According to an embodiment of the present invention, the prepayment account selection module 750 may include a prepayment account method selection module.

[0167] According to an embodiment of the present invention, the prepayment account method selection module may include a maximum balance selection unit, a minimum difference selection unit, or a highest weight selection unit.

[0168] The maximum balance selection unit may be used to traverse the bank accounts bound to the payment account, compare the balances, and select the bank account with the maximum balance. In one embodiment, the maximum balance selection unit may be used to perform the operation S610 described above, which will not be described in detail here.

[0169] The minimum difference selection unit can be used to traverse the bank accounts bound to the payment account, calculate the difference between the balance of each bank account and the predicted payment amount, and select the bank account with the smallest difference. In one embodiment, the minimum difference selection unit can be used to perform the operation S620 described above, which will not be repeated here.

[0170] The highest weight selection unit may be used to select the bank account with the highest selection weight according to the pre-set selection weight of each bank account bound to the payment account. In one embodiment, the highest weight selection unit may be used to perform the operation S630 described above, which will not be described in detail here.

[0171] According to an embodiment of the present invention, any multiple modules among the payment account acquisition module 710, the binding module 720, the prediction data acquisition module 730, the judgment module 740, the prepayment account selection module 750 and the automatic recharge module 760 can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the payment account acquisition module 710, the binding module 720, the prediction data acquisition module 730, the judgment module 740, the prepayment account selection module 750 and the automatic recharge module 760 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 a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation modes of software, hardware and firmware or in a proper combination of any of them. Alternatively, at least one of the payment account acquisition module 710, binding module 720, prediction data acquisition module 730, judgment module 740, prepayment account selection module 750 and automatic recharge module 760 can be at least partially implemented as a computer program module, which can perform corresponding functions when executed.

[0172] Figure 8A block diagram of an electronic device suitable for the method for automatic recharging of a payment account according to an embodiment of the present invention is schematically shown.

[0173] like Figure 8 As shown, the electronic device 800 according to an embodiment of the present invention includes a processor 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage part 808 into a random access memory (RAM) 803. The processor 801 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 801 may also include an onboard memory for caching purposes. The processor 801 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.

[0174] In RAM 803, various programs and data required for the operation of electronic device 800 are stored. Processor 801, ROM 802 and RAM 803 are connected to each other via bus 804. Processor 801 performs various operations of the method flow according to the embodiment of the present invention by executing the programs in ROM 802 and / or RAM 803. It should be noted that the program can also be stored in one or more memories other than ROM 802 and RAM 803. Processor 801 can also perform various operations of the method flow according to the embodiment of the present invention by executing the programs stored in the one or more memories.

[0175] According to an embodiment of the present invention, the electronic device 800 may further include an input / output (I / O) interface 805, which is also connected to the bus 804. The electronic device 800 may further include one or more of the following components connected to the I / O interface 805: an input portion 806 including a keyboard, a mouse, etc.; an output portion 807 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 808 including a hard disk, etc.; and a communication portion 809 including a network interface card such as a LAN card, a modem, etc. The communication portion 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the I / O interface 805 as needed. A removable medium 811, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 810 as needed, so that a computer program read therefrom is installed into the storage portion 808 as needed.

[0176] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiment; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present invention is implemented.

[0177] According to an embodiment of the present invention, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, an apparatus or a device. For example, according to an embodiment of the present invention, the computer-readable storage medium may include the ROM 802 and / or RAM 803 described above and / or one or more memories other than ROM 802 and RAM 803.

[0178] The embodiment of the present invention also includes a computer program product, which includes a computer program, and the computer program contains program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the method provided by the embodiment of the present invention.

[0179] The computer program executes the above functions defined in the system / device of the embodiment of the present invention when it is executed by the processor 801. According to the embodiment of the present invention, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0180] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 809, and / or installed from a removable medium 811. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0181] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 809, and / or installed from the removable medium 811. When the computer program is executed by the processor 801, the above functions defined in the system of the embodiment of the present invention are performed. According to the embodiment of the present invention, the system, device, means, module, unit, etc. described above can be implemented by a computer program module.

[0182] According to an embodiment of the present invention, the program code for executing the computer program provided by the embodiment of the present invention can be written in any combination of one or more programming languages, specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, such as Java, C++, python, "C" language or similar programming languages. The program code can be executed completely on the user computing device, partially on the user device, partially on the remote computing device, or completely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (e.g., using an Internet service provider to connect through the Internet).

[0183] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0184] The embodiments of the present invention are described above. However, these embodiments are only for the purpose of illustration, and are not intended to limit the scope of the present invention. Although each embodiment is described above, it does not mean that the measures in each embodiment cannot be used in combination advantageously. Without departing from the scope of the present invention, those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present invention.

Claims

1. A method for automatically recharging a payment account, characterized in that: The method comprises: Obtain payment account; Based on the payment account, binding a bank account corresponding to the payment account; Based on the federated prediction model, obtaining the predicted payment amount and predicted payment time of the payment account, wherein the federated prediction model is pre-trained based on the federated learning framework and using the historical payment information of the payment account as training data; Based on the predicted payment time, determining the timing of automatic recharge; In response to reaching the automatic recharge opportunity, selecting a prepaid account from the bank account bound to the payment account; and Based on the predicted payment amount, funds are transferred from the prepaid account to the payment account for automatic recharging.

2. The method according to claim 1, characterized in that Pre-training federated prediction models, including: Establishing a federated learning framework, the federated learning framework includes a server end and n clients, wherein the n clients include a bank end and a charging enterprise end, and the bank end and the charging enterprise end include respective historical payment information databases, wherein n is a positive integer greater than 1; Initializing basic models at the n clients respectively, and using data in the historical payment information database of each of the n clients as training data to perform local model training respectively, to obtain first model parameters of the n clients; Parameter aggregation operation: the n clients upload their respective first model parameters to the server, and the server aggregates the first model parameters of the n clients to obtain second model parameters; Model update operation: the server sends the second model parameters to the n clients, and the n clients respectively re-train local models based on the second model parameters and the historical payment information database, and update the first model parameters of the n clients; Parameter re-aggregation operation: the n clients upload the updated first model parameters to the server, and the server aggregates the updated first model parameters of the n clients to obtain updated second model parameters; and The parameter aggregation operation, the model updating operation and the parameter re-aggregation operation are repeated until the model converges or a maximum number of iterations is reached.

3. The method according to claim 2, characterized in that The server aggregates the first model parameters of the n clients in a weighted average manner, wherein the weights of the first model parameters of the n clients are determined according to the data amounts of the n clients.

4. The method according to claim 2, characterized in that: In the model updating operation, the first model parameters are updated using empirical risk minimization.

5. The method according to any one of claims 1 to 4, characterized in that: The step of binding a bank account corresponding to the payment account based on the payment account specifically includes: Obtaining identity authentication information of the payment account; Using an information matching method, matching the identity verification information with bank account information in a plurality of bank account databases; and In response to successful matching, the payment account is bound to the bank account.

6. The method according to claim 5, characterized in that The information matching method is used to match the identity authentication information with bank account information in multiple bank account databases, specifically including: Performing data preprocessing on the identity verification information; selecting matching criteria based on the data format of the bank account database; and A target matching algorithm is selected to search for matching accounts in a bank account database, wherein the target matching algorithm includes: string matching, edit distance algorithm or hash function.

7. The method according to claim 5, characterized in that: The selecting a prepayment account from the bank account corresponding to the payment account specifically includes: Traverse the bank accounts bound to the payment account, compare the balances, and select the bank account with the largest balance; or Traversing the bank accounts bound to the payment account, calculating the difference between the balance of each bank account and the predicted payment amount, and selecting the bank account with the smallest difference; or According to the pre-set selection weights of the bank accounts bound to each payment account, the bank account with the highest selection weight is selected.

8. A device for automatically recharging a payment account, characterized in that: The device comprises: The payment account acquisition module is used to: obtain the payment account; A binding module, used to: bind the bank account corresponding to the payment account based on the payment account; A prediction data acquisition module, used to: acquire the predicted payment amount and predicted payment time of the payment account based on a federated prediction model, wherein the federated prediction model is pre-trained based on a federated learning framework and using the historical payment information of the payment account as training data; A judgment module, used to: judge the timing of automatic recharge based on the predicted payment time; a prepayment account selection module, configured to: in response to reaching the automatic recharge timing, select a prepayment account from the bank account corresponding to the payment account; and The automatic recharge module is used to transfer funds from the prepaid account to the payment account based on the predicted payment amount to perform automatic recharge.

9. An electronic device, comprising: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors are enabled to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, causes the processor to execute the method according to any one of claims 1 to 7.

Images