Transactional document verification

Abstract

Disclosed are a transactional document verification method and apparatus, a storage medium, and an electronic device. The method comprises: an electronic device uploads a target model gradient parameter of a local transactional document verification model to a service platform, wherein the target model gradient parameter is used for instructing the service platform to determine a target model parameter on the basis of the target model gradient parameter uploaded by at least one electronic device; the electronic device receives the target model parameter sent by the service platform, and updates the local transactional document verification model on the basis of the target model parameter to obtain a target transactional document verification model; and the electronic device then performs transactional document verification processing on the basis of the target transactional document verification model.

Description

Transaction voucher review Technical Field

[0001] This specification relates to the field of computer technology, and in particular to transaction voucher auditing methods, devices, storage media, and electronic devices. Background Art

[0002] With the continuous development of the economy and the improvement of people's living standards, personal consumption demand has gradually increased, and consumer finance institutions have also gradually emerged, providing convenience for users' small loan needs. When using credit products provided by consumer finance institutions, users must abide by the repayment terms of the consumer finance institutions to maintain their good credit. If a user's credit report shows a delinquent record, it will have many negative consequences. If a user disputes the delinquent record, the user can modify the delinquent record by submitting a credit objection application. Users submitting credit objection applications need to provide relevant transaction records as evidence. The consumer finance institution will then review the evidence provided by the user to confirm its authenticity and decide whether to revoke the delinquent record.

[0003] Summary of the Invention

[0004] This specification provides a transaction credential review method, device, storage medium, and electronic device. The technical solution is as follows.

[0005] In a first aspect, this specification provides a transaction credential review method, applied to an electronic device, the method comprising:

[0006] Uploading the target model gradient parameters of the local transaction credential audit model to the service platform, wherein the target model gradient parameters are used to instruct the service platform to determine the target model parameters based on the target model gradient parameters uploaded by at least one electronic device;

[0007] receiving the target model parameters sent by the service platform, and updating the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model;

[0008] Transaction credential audit processing is performed based on the target transaction audit model.

[0009] In a second aspect, this specification provides a transaction credential review method, which is applied to a service platform, and the method includes:

[0010] receiving target model gradient parameters uploaded by at least one electronic device;

[0011] Determining target model parameters based on the target model gradient parameters uploaded by each of the electronic devices;

[0012] The target model parameters are sent to each of the electronic devices, and the target model parameters are used to instruct each of the electronic devices to update the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model, and perform transaction credential audit processing based on the target transaction credential audit model.

[0013] In a third aspect, this specification provides a transaction credential auditing device, applied to an electronic device, the device comprising:

[0014] a parameter uploading module, configured to upload target model gradient parameters of a local transaction credential audit model to a service platform, wherein the target model gradient parameters are used to instruct the service platform to determine target model parameters based on the target model gradient parameters uploaded by at least one electronic device;

[0015] A model updating module, configured to receive the target model parameters sent by the service platform, and update the local transaction voucher audit model based on the target model parameters to obtain a target transaction voucher audit model;

[0016] The data processing module is used to perform transaction voucher audit processing based on the target transaction audit model.

[0017] In a fourth aspect, this specification provides a transaction credential review device, which is applied to a service platform, and the device includes:

[0018] A data receiving module, configured to receive target model gradient parameters uploaded by at least one electronic device;

[0019] a data processing module, configured to determine target model parameters based on the target model gradient parameters uploaded by each of the electronic devices;

[0020] A data sending module is used to send the target model parameters to each of the electronic devices, and the target model parameters are used to instruct each of the electronic devices to update the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model, and perform transaction credential audit processing based on the target transaction credential audit model.

[0021] In a fifth aspect, this specification provides a computer storage medium having a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing the above-mentioned method steps.

[0022] In a sixth aspect, this specification provides a computer program product, wherein the computer program product stores at least one instruction, and the at least one instruction is loaded by a processor to execute the above-mentioned method steps.

[0023] In a seventh aspect, this specification provides an electronic device, which may include: a memory and a processor; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the memory and executing the above-mentioned method steps.

[0024] The beneficial effects brought about by the technical solution provided in this specification include at least the following:

[0025] In an embodiment of the present specification, an electronic device uploads a target model gradient parameter of a local transaction credential audit model to a service platform. The target model gradient parameter is used to instruct the service platform to determine a target model parameter based on the target model gradient parameter uploaded by at least one electronic device. The electronic device then receives the target model parameter sent by the service platform and updates the local transaction credential audit model according to the target model parameter to obtain a target transaction credential audit model. The electronic device then performs transaction credential audit processing based on the target transaction credential audit model. The target transaction credential audit model used by the electronic device for transaction credential audit processing in an embodiment of the present specification is obtained by updating the local model with the target model parameter obtained by processing and calculating the gradient data of multiple electronic devices by the service platform, that is, the target transaction credential audit model is obtained by adopting an integrated training model method based on multi-party data. Therefore, the model obtained by the integrated training model method is used to perform transaction credential audit processing, which improves the recognition effect of the transaction credential while ensuring the security of the transaction credential audit data. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0027] FIG1 is a schematic diagram of the system architecture of a transaction voucher audit method provided by an embodiment of this specification;

[0028] FIG2 is a flow chart of a transaction voucher review method provided in an embodiment of this specification;

[0029] FIG3 is a flow chart of another transaction voucher review method provided in an embodiment of this specification;

[0030] FIG4 is a schematic diagram of a time node provided by an embodiment of this specification;

[0031] FIG5 is another schematic diagram of time nodes provided in an embodiment of this specification;

[0032] FIG6 is a flow chart of a transaction voucher review method provided in an embodiment of this specification;

[0033] FIG7 is a schematic diagram of data interaction between a service platform and an electronic device provided in an embodiment of this specification;

[0034] FIG8 is a flow chart of another transaction voucher review method provided in an embodiment of this specification;

[0035] FIG9 is a schematic diagram of the structure of a transaction credential auditing device provided in an embodiment of this specification;

[0036] FIG10 is a schematic diagram of the structure of a transaction credential auditing device provided in an embodiment of this specification;

[0037] FIG11 is a schematic structural diagram of an electronic device provided in an embodiment of this specification;

[0038] FIG12 is a schematic diagram of the structure of a service platform provided in an embodiment of this specification. DETAILED DESCRIPTION

[0039] In order to make the invention objectives, features, and advantages of the embodiments of this specification more obvious and easy to understand, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the embodiments described are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of this specification.

[0040] In the description of this specification, it should be understood that the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. In the description of this specification, it should be noted that, unless otherwise expressly specified and limited, "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices. For those of ordinary skill in the art, the specific meanings of the above terms in this specification can be understood according to the specific circumstances. In addition, in the description of this specification, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0041] When a user files a credit dispute, the consumer finance institution's ability to identify the user's provided transaction vouchers is crucial for reversing overdue records. In related technologies, consumer finance institutions rely on voucher recognition models to determine whether the user's provided transaction vouchers are forged. The data of individual consumer finance institutions is private and prohibited from external disclosure. Therefore, individual consumer finance institutions rely solely on their own data to train their voucher recognition models. However, due to the relatively small amount of data available for model training, these voucher recognition models often exhibit poor performance.

[0042] In order to solve the above technical problems, this specification is described in detail below in conjunction with specific embodiments.

[0043] Please refer to Figure 1, which is a schematic diagram of the system architecture of a transaction voucher review method provided in an embodiment of this specification.

[0044] As shown in FIG1 , FIG1 includes a service platform and electronic devices. The number of electronic devices and the number of nodes in the service platform shown in FIG1 are only exemplary, and the embodiments of this specification do not limit their number.

[0045] In the embodiments of this specification, the service platform may refer to a server consisting of one node or a server cluster consisting of multiple nodes.

[0046] When the service platform is a server cluster composed of multiple nodes, each node in the service platform can be a separate server device, such as: rack-mounted, blade, tower, or cabinet-type server equipment, or a workstation, mainframe computer and other hardware devices with strong computing capabilities; it can also be a server cluster composed of multiple servers. The servers in the service cluster can be composed in a symmetrical manner, where each server has equivalent functions and status in the transaction link, and each server can provide services to the outside world independently. Providing services to the outside world independently can be understood as not requiring the assistance of other servers.

[0047] In the embodiments of this specification, the electronic device may be a computer device having functions such as transaction credential review.

[0048] It should be noted that the electronic device and at least one node in the service platform establish a communication connection through a network for interactive communication. The network can be a wireless network or a wired network. The wireless network includes but is not limited to a cellular network, a wireless local area network, an infrared network or a Bluetooth network, and the wired network includes but is not limited to Ethernet, a universal serial bus (USB) or a controller area network. In one or more embodiments of the specification, technologies and / or formats including Hypertext Markup Language (HTML) and Extensible Markup Language (XML) are used to represent data exchanged over the network (such as a target compressed package). In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), and Internet Protocol Security (IPsec) can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above-mentioned data communication technologies.

[0049] In an embodiment of the present specification, an electronic device uploads a target model gradient parameter of a local transaction credential audit model to a service platform. The target model gradient parameter is used to instruct the service platform to determine a target model parameter based on the target model gradient parameter uploaded by at least one electronic device. The electronic device then receives the target model parameter sent by the service platform and updates the local transaction credential audit model according to the target model parameter to obtain a target transaction credential audit model. The electronic device then performs transaction credential audit processing based on the target transaction credential audit model. The target transaction credential audit model used by the electronic device for transaction credential audit processing in an embodiment of the present specification is obtained by updating the local model with the target model parameter obtained by processing and calculating the gradient data of multiple electronic devices by the service platform, that is, the target transaction credential audit model is obtained by adopting an integrated training model method based on multi-party data. Therefore, the model obtained by the integrated training model method is used to perform transaction credential audit processing, which improves the recognition effect of the transaction credential while ensuring the security of the transaction credential audit data.

[0050] Please refer to Figure 2, which is a flowchart of a transaction voucher review method provided in an embodiment of this specification. The execution subject of the method in this embodiment is an electronic device. As shown in Figure 2, the method in this embodiment of this specification may include the following steps S202 to S206.

[0051] S202, uploading the target model gradient parameters of the local transaction voucher audit model to the service platform.

[0052] As is easy to understand, the target model gradient parameter is used to instruct the service platform to determine the target model parameter based on the target model gradient parameter uploaded by at least one electronic device. The target model gradient parameter refers to the gradient of the model parameters of the local transaction credential audit model. The target model gradient parameter can be obtained by training the local transaction credential audit model using training data.

[0053] The local transaction voucher review model refers to a model that has the ability to review transaction vouchers based on a machine learning model. The local transaction voucher review model has the ability to review transaction vouchers, which means that the local transaction voucher review model has the ability to identify the authenticity of transaction review voucher data. Transaction review voucher data refers to the voucher data provided by the user during the application process when the user submits a transaction review application to a consumer finance institution. Specifically, in the embodiment of the present application, the scenario in which the user submits a transaction review application to a consumer finance institution may refer to a scenario in which the user submits a credit objection application to the overdue credit record given by the consumer finance institution. In this scenario, the transaction review voucher data may refer to voucher data such as the user's personal identification document and transaction record materials.

[0054] The local transaction credential audit model may have local model parameters, which are generated by the service platform and sent to the electronic device. The local transaction credential audit model may be obtained by the electronic device by updating the initial credential audit model based on the local model parameters. The initial credential audit model may have the same model architecture as a preset transaction credential audit model in the service platform, which has local model parameters.

[0055] The service platform refers to the server that, together with electronic devices, forms a federated learning architecture. The federated learning architecture refers to the system architecture shown in Figure 1. Specifically, the federated learning architecture includes a service platform and multiple electronic devices. In this architecture, electronic devices send gradient data to the service platform, which aggregates the gradient data sent by each electronic device and then sends the aggregated results to each electronic device.

[0056] In some embodiments, executing the step of S202 may specifically include: encrypting the target model gradient parameters of the local transaction credential model to obtain encrypted gradient data, and uploading the encrypted gradient data to the service platform. The target model gradient parameters are encrypted to obtain encrypted gradient data, and the encryption key used can be a preset encryption key agreed upon in advance by the electronic device and the service platform in a trusted execution environment, or it can be the public key of the service platform. The electronic device encrypts the target model gradient parameters, and accordingly, the service platform needs to decrypt the encrypted gradient data. When the encryption key is the above-mentioned preset encryption key, the decryption key used by the service platform is the preset decryption key corresponding to the preset encryption key; when the encryption key is the public key of the service platform, the decryption key used by the service platform is the private key of the service platform. In this way, by encrypting the target model gradient parameters and sending them to the service platform, the target model gradient parameters are avoided from being leaked and causing insecurity.

[0057] Optionally, in addition to encrypting and sending the target model gradient parameters to the service platform, encrypted gradient data obtained by encrypting the target model gradient parameters can be uploaded to the service platform when the current time reaches a preset sending time. For example, the preset sending time can be set to 9:00 PM or 10:00 PM every day.

[0058] S204: Receive target model parameters sent by the service platform, and update the local transaction voucher audit model based on the target model parameters to obtain a target transaction voucher audit model.

[0059] It is easy to understand that the target model parameters refer to the latest model parameters determined by the service platform based on the processed data obtained by processing the target model gradient parameters sent by each electronic device.

[0060] In some embodiments, receiving the target model parameters sent by the service platform may involve the service platform proactively sending the target model parameters to the electronic device, allowing the electronic device to receive the target model parameters. Alternatively, the electronic device may send a parameter query request to the service platform, and the service platform, based on the parameter query request, sends the target model parameters to the electronic device, allowing the electronic device to receive the target model parameters. Specifically, receiving the target model parameters sent by the service platform by the electronic device may include: the electronic device receiving encrypted parameter data corresponding to the target model parameters sent by the service platform, and the electronic device decrypting the encrypted parameter data to obtain the target model parameters. The service platform encrypts the target model parameters to obtain the encrypted parameter data, and the encryption key used may be a preset encryption key agreed upon between the electronic device and the service platform in a trusted execution environment, or may be the public key of the electronic device. Accordingly, the service platform encrypts the target model parameters, and the electronic device needs to decrypt the encrypted parameter data. When the encryption key is the preset encryption key, the decryption key used by the electronic device is the preset decryption key corresponding to the preset encryption key; when the encryption key is the public key of the electronic device, the decryption key used by the electronic device is the private key of the electronic device. In this way, by receiving the encrypted data of the target model gradient parameters sent by the service platform, the target model parameters are prevented from being leaked and thus becoming insecure.

[0061] The local transaction voucher audit model is updated based on the target model parameters to obtain the target transaction voucher audit model. Specifically, the local model parameters of the local transaction voucher audit model are updated to the target model parameters to obtain the target transaction voucher audit model.

[0062] S206: Perform transaction voucher audit processing based on the target transaction audit model.

[0063] In some embodiments, the target transaction audit voucher data of the target user can be obtained first, and then the target transaction audit model can be used to perform transaction voucher audit processing on the target transaction audit voucher data to obtain the transaction voucher audit result corresponding to the target user. The target transaction audit voucher data may include the target user's personal identification document, transaction record materials and other voucher data; the target transaction audit voucher data may be uploaded directly to the electronic device by the user using the user terminal, or it may be forwarded to the electronic device by other electronic devices belonging to the same consumer finance structure as the electronic device. The transaction voucher audit result may include information such as the fraud score and fraud information corresponding to the target user.

[0064] Optionally, after obtaining the transaction credential review result corresponding to the target user, the electronic device may also send the transaction credential review result to the user terminal so that the user terminal can know the review result of the credential data.

[0065] Optionally, steps S202-S204 of the embodiments of this specification may be executed in a loop, where the loop terminates when the target model parameters received by the electronic device from the service platform are the model parameters indicated by the service platform at the time of model convergence. In other words, when the target model parameters received by the electronic device are the model parameters indicated by the service platform at the time of model convergence, the electronic device will no longer execute steps S202-S204 during the period in which it does not receive the gradient calculation instruction. Instead, after obtaining the transaction audit credential data of the target user, the electronic device will use the target transaction credential audit model obtained at the time of model convergence to perform transaction credential audit processing.

[0066] In an embodiment of the present specification, an electronic device uploads a target model gradient parameter of a local transaction credential audit model to a service platform. The target model gradient parameter is used to instruct the service platform to determine a target model parameter based on the target model gradient parameter uploaded by at least one electronic device. The electronic device then receives the target model parameter sent by the service platform and updates the local transaction credential audit model according to the target model parameter to obtain a target transaction credential audit model. The electronic device then performs transaction credential audit processing based on the target transaction credential audit model. The target transaction credential audit model used by the electronic device for transaction credential audit processing in an embodiment of the present specification is obtained by updating the local model with the target model parameter obtained by processing and calculating the gradient data of multiple electronic devices by the service platform, that is, the target transaction credential audit model is obtained by adopting an integrated training model method based on multi-party data. Therefore, the model obtained by the integrated training model method is used to perform transaction credential audit processing, which improves the recognition effect of the transaction credential while ensuring the security of the transaction credential audit data.

[0067] Please refer to Figure 3, which is a flowchart of a transaction voucher review method provided in an embodiment of this specification. The execution subject of the method in this embodiment is an electronic device. As shown in Figure 3, the method in this embodiment of this specification may include the following steps S302 to S310.

[0068] S302: Determine a local transaction credential audit model and obtain target model gradient parameters of the local transaction credential audit model.

[0069] In some embodiments, executing the step of determining the local transaction credential audit model may specifically include: obtaining multiple first transaction credential audit models, and determining the local transaction credential audit model from the multiple first transaction credential audit models based on the generation times corresponding to the multiple first transaction credential audit models. Specifically, the first transaction credential audit model corresponding to the generation time closest to the current time may be used as the local transaction credential audit model. For example, the electronic device may receive the model parameters sent by the service platform every day. Assuming that a certain day is today, the electronic device may use the model parameters received today to update the first transaction credential audit model generated on the previous day to obtain a new first transaction credential audit model. The transaction credential recognition capability of each new first transaction credential audit model is better than the transaction credential recognition capability of the previous first transaction credential audit model.

[0070] The steps of obtaining the target model gradient parameters of the local transaction credential audit model may specifically include: A2: obtaining historical transaction audit credential data of historical users; A4: performing model training processing on the local transaction credential audit model based on the historical transaction audit credential data to obtain the target model gradient parameters of the local transaction credential audit model.

[0071] Among them, the historical transaction audit voucher data may be the transaction audit voucher data uploaded by the user terminal received by the electronic device within a preset time period; in this scenario, the electronic device may communicate directly with the user terminal corresponding to the user of the consumer finance institution. In another scenario, the historical transaction audit voucher data may be the transaction audit voucher data sent by the first electronic device received by the electronic device, and the transaction audit voucher data may be data obtained by the first electronic device within a preset time period; at this time, the electronic device and the first electronic device may be computer devices belonging to the same consumer finance institution, and the first electronic device may be used to communicate with the user terminal corresponding to the user of the consumer finance institution to obtain the transaction audit voucher data, and forward the transaction audit voucher data to the electronic device, and the electronic device may be used to process the transaction audit voucher data. The specific interpretation of the transaction audit voucher data can be found in the description of S202 in the embodiment shown in Figure 2, and will not be repeated here.

[0072] Please refer to the time node schematic diagram shown in Figure 4. The preset time period is explained in conjunction with Figure 4 below: As shown in Figure 4, assuming that the generation time of the local transaction voucher audit model is the first time of the day, and the time when the first transaction voucher audit model is generated the day before is the first time of the day before, the preset time period may refer to a time period starting from the first time of the day before and ending at the first time of the day.

[0073] When executing step A4, specifically, if it is determined that the historical transaction audit voucher data meets the preset model training conditions, then model training processing is performed on the local transaction audit voucher model based on the historical transaction audit voucher data to obtain the target model gradient parameters of the local transaction voucher audit model. Determining that the historical transaction audit voucher data meets the preset model training conditions can be implemented in two specific ways: first, determining whether the cumulative amount of historical transaction audit voucher data is greater than the preset training amount; if so, determining that the historical transaction audit voucher data meets the preset model training conditions; second, determining whether the acquisition time of the historical transaction audit voucher data is within the preset training time period; if so, determining that the historical transaction audit voucher data meets the preset model conditions.

[0074] The cumulative number refers to the total number of historical transaction audit voucher data obtained during the above-mentioned preset time period. The preset training time period may be the same as the above-mentioned preset time period, or the time range of the preset training time period may be smaller than the time range of the above-mentioned preset time period.

[0075] Specifically, the local transaction voucher audit model is trained based on the historical transaction voucher data to obtain the target model gradient parameters of the local transaction voucher audit model. This can be understood as inputting a number of historical transaction voucher audit data as batch training data into the local transaction voucher audit model for forward propagation calculation to obtain the output of the local transaction voucher audit model, and then using backward propagation to calculate the gradients of each model of the local transaction voucher audit model on the batch training data based on the output of the local transaction voucher audit model, which is the target model gradient parameter.

[0076] S304: Upload the target model gradient parameters of the local transaction voucher audit model to the service platform.

[0077] S306: Receive target model parameters sent by the service platform, and update the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model.

[0078] Specifically, the implementation of S304 and S306 can be specifically referred to the explanation of S202 and S204 in the embodiment shown in FIG2 , which will not be repeated here.

[0079] S308, obtaining the transaction audit voucher data of the target user.

[0080] In the embodiments of this specification, the transaction audit credential data of the target user may be data obtained within a first preset time period. Specifically, the transaction audit credential data of the target user may be transaction audit credential data uploaded by a user terminal and received by the electronic device within the first preset time period; or it may be transaction audit credential data sent by a first electronic device and received by the electronic device. The transaction audit credential data may be transaction audit credential data sent by a user terminal and received by the first electronic device within the first preset time period. The first electronic device and the electronic device may be computer devices belonging to the same consumer finance institution.

[0081] The first preset time period may refer to the time period after the target transaction voucher audit model is generated. Referring to the time node diagram shown in FIG5 , the first preset time period is explained below in conjunction with FIG5 : As shown in FIG5 , assuming that the target transaction voucher audit model is generated at the first time of a certain day, and the local transaction voucher audit model is generated at the first time of the day before the certain day, the first preset time period may refer to the time period starting at the first time of the certain day and ending at the second time of the certain day.

[0082] S310, performing transaction voucher audit processing on the transaction audit voucher data based on the target transaction voucher audit model to obtain a transaction voucher audit result corresponding to the target user.

[0083] Specifically, the steps of executing S310 may include: B2: performing voucher classification processing based on the transaction audit voucher data to obtain target transaction audit voucher data, where the target transaction audit voucher data carries the target voucher classification corresponding to the transaction audit voucher data; B4: performing transaction voucher audit processing on the target transaction audit voucher data based on the target transaction voucher audit model to obtain the transaction voucher audit result corresponding to the target user.

[0084] The transaction voucher audit result may include information such as a fraud score, fraud basis, and fraud information for the transaction audit voucher data. Optionally, the greater the fraud score, the greater the probability of the target user committing fraud.

[0085] When executing step B2, the following steps may be performed: performing voucher classification and identification processing on the transaction audit voucher data to obtain a target voucher classification corresponding to the transaction audit voucher data; and labeling the transaction audit voucher data based on the target voucher classification to obtain target transaction audit voucher data. Transaction audit voucher data may include voucher data such as a user's personal identification document, transaction record materials, error record statements, and medical certification materials. Accordingly, the target voucher classification may include categories such as identification documents, transaction records, personal statements, and medical materials. The transaction audit voucher data may first be subjected to voucher classification and identification processing to obtain a target voucher classification for the various voucher data contained in the transaction audit voucher data. The target voucher classification may then be added to the various voucher data contained in the transaction audit voucher data in the form of a classification label to obtain target transaction audit voucher data containing the target voucher classification. In this manner, by performing transaction voucher audit processing based on voucher data containing accurate voucher classifications, an accurate transaction voucher audit result identified by the target transaction audit voucher model may be obtained.

[0086] When executing step B2, the transaction audit voucher data may be input into a voucher classification and recognition model to obtain target transaction audit voucher data. The voucher classification and recognition model is trained using sample target transaction audit voucher data labels corresponding to labeled sample transaction audit voucher data. The following explains the model training process of the voucher classification and recognition model.

[0087] Model creation: Create an initial credential classification recognition model for credential classification scenarios based on a machine learning model. Sample data acquisition: Acquire a large amount of sample data. The sample data is based on the historical transaction audit credential data acquired by the electronic device to process and extract historical credential classifications to generate sample data containing historical credential classifications. Sample data labeling: Based on the needs of the credential classification scenario, an expert service is introduced to manually label the sample data with corresponding sample labels. The sample labels include credential classification labels for each sample data. Model training process: The sample data is input into the initial credential classification recognition model for at least one round of model training to obtain a predicted credential classification. The model loss value is determined based on the predicted credential classification and the sample data label (credential classification label). The model parameters of the initial credential classification recognition model are adjusted based on the model loss value until the model training end conditions are met to obtain a credential classification recognition model.

[0088] Optionally, the model training termination conditions of the model may include, for example, the value of the loss function is less than or equal to a preset loss function threshold, the number of iterations reaches a preset number threshold, etc. The specific model training termination conditions can be determined based on actual conditions and are not specifically limited here.

[0089] It should be noted that the machine learning models involved in one or more embodiments of this specification include but are not limited to the fitting of one or more machine learning models such as convolutional neural network (CNN) model, deep neural network (DNN) model, recurrent neural network (RNN) model, embedding model, gradient boosting decision tree (GBDT) model, logistic regression (LR) model, etc.

[0090] Optionally, steps S302-S306 of the embodiments of this specification may be executed in a loop, where the loop terminates when the target model parameters received by the electronic device from the service platform are the model parameters indicated by the service platform at the time of model convergence. In other words, when the target model parameters received by the electronic device are the model parameters indicated by the service platform at the time of model convergence, the electronic device will no longer execute steps S302-S306 during the period in which it does not receive the gradient calculation instruction. Instead, after obtaining the transaction audit credential data of the target user, the electronic device will perform transaction credential audit processing on the transaction audit credential data using the target transaction credential audit model obtained at the time of model convergence.

[0091] In an embodiment of the present specification, the electronic device avoids sending target model gradient parameters with error risks to the service platform by first determining the local transaction credential audit model and then determining the target model gradient parameters. After receiving the target model parameters sent by the service for the target model gradient parameters, the local transaction credential audit model is updated by using the target model parameters to obtain a target transaction credential audit model with higher recognition capability, so as to achieve the effect of improving the accuracy of the transaction credential audit results corresponding to the target user when the transaction audit credential data of the target user is identified.

[0092] Please refer to Figure 6, which is a flowchart of a transaction voucher review method provided in an embodiment of this specification. The execution subject of the method in this embodiment is a service platform. As shown in Figure 6, the method in this embodiment of this specification may include the following steps S602 to S606.

[0093] S602: Receive target model gradient parameters uploaded by at least one electronic device.

[0094] In some embodiments, executing step S602 may specifically include: receiving encrypted gradient data uploaded by at least one electronic device, decrypting each encrypted gradient data to obtain target model gradient parameters corresponding to each electronic device. The encrypted gradient data sent by different electronic devices may be decrypted using different decryption keys. For each electronic device, the encrypted gradient data is decrypted to obtain the target model gradient parameters. The decryption key used may be a preset decryption key agreed upon between the electronic device and the service platform in a trusted execution environment, or the public key of the electronic device. The service platform decrypts the encrypted gradient data, and accordingly, the electronic device is required to encrypt the target model gradient parameters. When the decryption key used by the service platform is the preset decryption key, the encryption key used by the electronic device may be a preset encryption key corresponding to the preset decryption key. When the decryption key used by the service platform is the public key of the electronic device, the encryption key used by the electronic device may be the private key of the electronic device. In this way, the service platform receives the encrypted target model gradient parameters sent by the electronic device, preventing the target model gradient parameters from being leaked and potentially insecure.

[0095] Optionally, before executing S602, the service platform may share the local transaction credential audit model with each electronic device. Specifically, the service platform may send local model parameters of the local transaction credential audit model to the electronic device. The local model parameters may be the initial parameters of the local transaction credential audit model. The local transaction credential audit model may be a model trained by the service platform based on sample credential data.

[0096] S604: Determine target model parameters based on target model gradient parameters uploaded by each electronic device.

[0097] The target model parameters refer to the latest model parameters determined by the service platform based on the processed data obtained by processing the target model gradient parameters sent by each electronic device. In the embodiment of the present application, the target model parameters refer to the latest model parameters of the local transaction credential audit model shared by the service platform and each electronic device.

[0098] In some embodiments, the service platform may perform gradient aggregation processing on the target model gradient parameters of the same batch to obtain an aggregated gradient, and update the local model parameters of the local transaction credential audit model based on the aggregated gradient to obtain the target model parameters. The target model gradient parameters of the same batch may refer to the target model gradient parameters received within the same time period. The service platform and each electronic device may agree in advance on the time period for uploading the target model gradient parameters in each round. In the process of calculating the target model parameters in each round, the target model gradient parameters within this time period may be used for calculation to obtain the target model parameters.

[0099] S606: Send the target model parameters to each electronic device.

[0100] The target model parameters are used to instruct each electronic device to update the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model, and perform transaction credential audit processing based on the target transaction credential audit model.

[0101] In some embodiments, executing step S606 may specifically include: encrypting the target model parameters to obtain encrypted parameter data, and sending the encrypted parameter data to each electronic device. The encryption keys used to encrypt the encrypted parameter data sent to different electronic devices may be different. For each target model parameter sent to each electronic device, the target model parameters are encrypted to obtain encrypted parameter data. The encryption key used may be a preset encryption key agreed upon between the electronic device and the service platform in a trusted execution environment, or may be the public key of the electronic device. The service platform encrypts the target model parameters, and accordingly, the electronic device is required to decrypt the encrypted parameter data. When the encryption key used by the service platform is the preset encryption key, the key used by the electronic device for decryption may be a preset decryption key corresponding to the preset encryption key; when the encryption key used by the service platform is the public key of the electronic device, the key used by the electronic device for decryption may be the private key of the electronic device. In this way, the service platform sends the encrypted target model parameters to the electronic device, preventing the target model parameters from being leaked and potentially insecure.

[0102] It should be noted that the service platform can calculate multiple rounds of target model parameters using multiple batches of target model gradient parameters to continuously update the local transaction credential audit model shared with each electronic device until the local transaction credential audit model converges. Each time the service platform updates the local transaction credential audit model, it can send the latest model parameters of the local transaction credential audit model, i.e., the target model parameters, to each electronic device.

[0103] Please refer to the data interaction diagram between the service platform and the electronic device shown in Figure 7. The following is an explanation of the process of each update of the local transaction credential audit model by the service platform in conjunction with Figure 7: Step 1, the service platform sends the first model parameters to each electronic device; Step 2, each electronic device updates the local transaction credential audit model based on the first model parameters to obtain the first transaction credential audit model, and each electronic device uploads the first model gradient parameters of the first transaction credential audit model to the service platform; Step 3, the service platform determines the second model parameters based on the first model gradient parameters, and the service platform sends each second model parameter to each electronic device. The above steps 1 to 2 are repeated to achieve the purpose of jointly training the transaction credential audit model between the service platform and each electronic device until the transaction credential audit model converges.

[0104] In the embodiments of this specification, the service platform receives target model gradient parameters uploaded by each electronic device and determines target model parameters based on each target model gradient parameter, thereby achieving the effect of integrating and training the model using multi-party data, thereby improving the model's recognition capabilities. The service platform then sends the target model parameters to each electronic device, allowing the electronic device to update the local transaction credential review model shared by the service platform based on the target model parameters to obtain a target transaction credential review model with better recognition capabilities, thereby improving the electronic device's recognition of transaction credentials using the target transaction credential review model.

[0105] Please refer to Figure 8, which is a flowchart of a transaction voucher review method provided in an embodiment of this specification. The execution subject of the method in this embodiment is a service platform. As shown in Figure 7, the method in this embodiment of this specification may include the following steps S802 to S808.

[0106] S802: Receive target model gradient parameters uploaded by at least one electronic device.

[0107] Specifically, please refer to the description of S602 in the embodiment shown in FIG6 , which will not be repeated here.

[0108] S804: Determine the target model gradient mean based on the target model gradient parameters uploaded by each electronic device.

[0109] In some embodiments, when executing S804, it may specifically include: obtaining the number of samples corresponding to the target model gradient parameter, determining the target weight corresponding to the target model gradient parameter based on the number of samples; and calculating the target model gradient mean based on the target model gradient parameter and the target weight.

[0110] Among them, the number of samples refers to the number of sample transaction audit voucher data used in the process of the electronic device calculating the target model gradient parameters. The sample transaction audit voucher data is the training data used in the process of the electronic device training the local transaction voucher audit model to obtain the target model gradient parameters.

[0111] The service platform can be configured with a preset weight mapping relationship between the number of reference samples and the reference weight. For the number of samples corresponding to the target model gradient parameters sent by different electronic devices, the target weight corresponding to the number of samples can be queried based on the above weight mapping relationship.

[0112] For example, there are three electronic devices, and the target model gradient parameters of these electronic devices are g1, g2, and g3 respectively; the target weight determined according to the number of samples corresponding to g1 is m1, the target weight determined according to the number of samples corresponding to g2 is m2, and the target weight determined according to the number of samples corresponding to g3 is m3; the target model gradient mean g0 = g1*m1+g2*m2+g3*m3 can be obtained.

[0113] The embodiment of the present application calculates the gradient mean to comprehensively evaluate the gradient parameters of each target model, thereby ensuring the accuracy of the gradient used to update the model parameters.

[0114] S806: Update the local model parameters based on the target model gradient mean to obtain the target model parameters.

[0115] It is easy to understand that the local model parameters refer to the model parameters of the latest version of the local transaction credential audit model shared by the service platform and each electronic device.

[0116] In some embodiments, the target model gradient mean and the local model parameters may be calculated according to a preset parameter update formula to obtain the target model parameters.

[0117] For example, the preset parameter update formula may be w=w0-a*g0, where w is the target model parameter, w0 is the local model parameter, a is the learning rate of the model parameter, and g0 is the target model gradient mean.

[0118] S808: Send the target model parameters to each electronic device.

[0119] For details, please refer to the description of S606 in the embodiment shown in FIG6 , which will not be repeated here.

[0120] In the embodiment of this specification, the service platform receives the target model gradient parameters uploaded by each electronic device to determine the target model gradient mean based on each target model gradient parameter, and then determines the target model parameters based on the target model gradient mean. While achieving the effect of integrating the training model using multi-party data, it also achieves the effect of improving the accuracy of the gradient by calculating the target model gradient mean, and then uses the more accurate gradient to calculate the more accurate target model parameters, and uses the more accurate target model parameters to update the model, thereby better improving the recognition ability of the model. The service platform then sends the more accurate target model parameters to each electronic device, which enables the electronic device to update the local transaction voucher audit model shared by the service platform based on the more accurate target model parameters to obtain a target transaction voucher audit model with better recognition ability, thereby better improving the electronic device's recognition effect on transaction vouchers using the target transaction voucher audit model.

[0121] The following will describe in detail the transaction voucher auditing device provided by the embodiment of the present application in conjunction with Figure 9. It should be noted that the transaction voucher auditing device shown in Figure 9 is used to execute the method of the embodiment shown in Figures 2 and 3 of the present application. For ease of explanation, only the parts related to the embodiment of this specification are shown. For specific technical details not disclosed, please refer to the embodiment shown in Figures 2 and 3 of the present application.

[0122] Please refer to Figure 9, which shows a schematic diagram of the structure of a transaction credential auditing device according to an embodiment of this specification. The transaction credential auditing device 900 can be implemented as all or part of a device through software, hardware, or a combination of both. According to some embodiments, the transaction credential auditing device 1 includes a parameter upload module 911, a model update module 912, and a data processing module 913.

[0123] The parameter upload module 911 is used to upload the target model gradient parameters of the local transaction credential audit model to the service platform, and the target model gradient parameters are used to instruct the service platform to determine the target model parameters based on the target model gradient parameters uploaded by at least one electronic device.

[0124] The model updating module 912 is configured to receive the target model parameters sent by the service platform, and update the local transaction voucher audit model based on the target model parameters to obtain a target transaction voucher audit model.

[0125] The data processing module 913 is used to perform transaction voucher audit processing based on the target transaction audit model.

[0126] Optionally, the data processing module 913 includes:

[0127] The first processing unit is used to obtain transaction audit voucher data of a target user;

[0128] The second processing unit is configured to perform transaction voucher audit processing on the transaction audit voucher data based on the target transaction voucher audit model to obtain a transaction voucher audit result corresponding to the target user.

[0129] Optionally, the second processing unit includes:

[0130] A first subunit is configured to perform voucher classification processing based on the transaction audit voucher data to obtain target transaction audit voucher data, wherein the target transaction audit voucher data carries a target voucher classification corresponding to the transaction audit voucher data;

[0131] The second subunit is configured to perform transaction voucher audit processing on the target transaction audit voucher data based on the target transaction voucher audit model to obtain a transaction voucher audit result corresponding to the target user.

[0132] Optionally, the second subunit is specifically configured to:

[0133] Perform voucher classification identification processing on the transaction audit voucher data to obtain a target voucher classification corresponding to the transaction audit voucher data;

[0134] The transaction audit voucher data is labeled based on the target voucher classification to obtain target transaction audit voucher data.

[0135] Optionally, the second subunit is specifically configured to:

[0136] The transaction audit voucher data is input into a voucher classification and recognition model to obtain target transaction audit voucher data. The voucher classification and recognition model is trained with sample target transaction audit voucher data labels corresponding to labeled sample transaction audit voucher data.

[0137] Optionally, the transaction voucher auditing device 1 further includes:

[0138] The parameter determination module is used to determine a local transaction credential audit model and obtain target model gradient parameters of the local transaction credential audit model.

[0139] Optionally, the parameter determination module includes:

[0140] Parameter acquisition module, used to obtain historical transaction audit certificate data of historical users;

[0141] A parameter generation module is used to perform model training processing on the local transaction credential audit model based on the historical transaction audit credential data to obtain the target model gradient parameters of the local transaction credential audit model.

[0142] Optionally, the parameter generation module includes:

[0143] A parameter generation unit is used to perform model training processing on the local transaction voucher audit model based on the historical transaction audit voucher data if it is determined that the historical transaction audit voucher data meets the preset model training conditions, so as to obtain the target model gradient parameters of the local transaction voucher audit model.

[0144] Optionally, the parameter generation unit is specifically configured to:

[0145] Determining whether the accumulated amount of the historical transaction audit voucher data is greater than a preset training amount, and if the accumulated amount is greater than the preset training amount, determining that the historical transaction audit voucher data meets the preset model training condition; or,

[0146] It is determined whether the acquisition time of the historical transaction audit voucher data is within a preset training time period. If the acquisition time is within the preset training time period, it is determined that the historical transaction audit voucher data meets the preset model condition.

[0147] The transaction voucher auditing device provided in the embodiment of the present application will be described in detail below in conjunction with Figure 10. It should be noted that the transaction voucher auditing device shown in Figure 10 is used to execute the method of the embodiment shown in Figures 6 and 8 of the present application. For ease of explanation, only the parts related to the embodiment of this specification are shown. For specific technical details not disclosed, please refer to the embodiment shown in Figures 6 and 8 of the present application.

[0148] Please refer to Figure 10, which shows a schematic diagram of the structure of a transaction credential auditing device according to an embodiment of this specification. The transaction credential auditing device 1000 can be implemented as all or part of a device through software, hardware, or a combination of both. According to some embodiments, the transaction credential auditing device 1 includes a data receiving module 1011, a data processing module 1012, and a data sending module 1013.

[0149] The data receiving module 1011 is configured to receive target model gradient parameters uploaded by at least one electronic device.

[0150] The data processing module 1012 is configured to determine target model parameters based on the target model gradient parameters uploaded by each of the electronic devices.

[0151] The data sending module 1013 is used to send the target model parameters to each of the electronic devices. The target model parameters are used to instruct each of the electronic devices to update the local transaction credential audit model based on the target model parameters to obtain the target transaction credential audit model, and perform transaction credential audit processing based on the target transaction credential audit model.

[0152] Optionally, the data processing module 1012 includes:

[0153] a first calculation unit, configured to determine a target model gradient mean based on the target model gradient parameters uploaded by each of the electronic devices;

[0154] The second computing unit is configured to update the local model parameters based on the target model gradient mean to obtain the target model parameters.

[0155] Optionally, the first computing unit is specifically configured to:

[0156] Obtaining the number of samples corresponding to the target model gradient parameter, and determining the target weight corresponding to the target model gradient parameter based on the number of samples;

[0157] A target model gradient mean is calculated based on the target model gradient parameter and the target weight.

[0158] Please refer to Figure 11, which shows a schematic diagram of the structure of an electronic device provided in accordance with an exemplary embodiment of this specification. The electronic device described herein may include one or more of the following components: a processor 110, a memory 120, an input device 130, an output device 140, and a bus 150. The processor 110, the memory 120, the input device 130, and the output device 140 may be connected via the bus 150.

[0159] The processor 110 may include one or more processing cores. The processor 110 utilizes various interfaces and circuits to connect various components within the terminal. It executes instructions, programs, code sets, or instruction sets stored in the memory 120, as well as accesses data stored in the memory 120, to perform various functions of the terminal 100 and process data. Optionally, the processor 110 may be implemented using at least one of the following hardware forms: a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). The processor 110 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem handles wireless communications. It is understood that the modem may not be integrated into the processor 110 and may be implemented separately via a communications chip.

[0160] The memory 120 may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory 120 includes a non-transitory computer-readable storage medium. The memory 120 may be used to store instructions, programs, codes, code sets, or instruction sets. The memory 120 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc. The operating system may be an Android system, including a system deeply developed based on the Android system, an iOS system developed by Apple, including a system deeply developed based on the iOS system, or other systems.

[0161] In order for the operating system to distinguish the specific application scenarios of third-party applications, it is necessary to open up data communication between third-party applications and the operating system so that the operating system can obtain the current scenario information of third-party applications at any time, and then perform targeted system resource adaptation based on the current scenario.

[0162] The input device 130 is used to receive input commands or data and includes, but is not limited to, a keyboard, a mouse, a camera, a microphone, or a touch-sensitive device. The output device 140 is used to output commands or data and includes, but is not limited to, a display device and a speaker. In one example, the input device 130 and the output device 140 may be combined, and the input device 130 and the output device 140 may be a touch-sensitive display.

[0163] The touch display screen can be designed as a full screen, a curved screen or a special-shaped screen. The touch display screen can also be designed as a combination of a full screen and a curved screen, or a combination of a special-shaped screen and a curved screen, which is not limited in the embodiments of this specification.

[0164] In addition, those skilled in the art will understand that the structures of the electronic devices shown in the above figures do not limit the electronic devices. The electronic devices may include more or fewer components than shown, or may combine certain components, or arrange the components differently. For example, the electronic devices may also include radio frequency circuits, input units, sensors, audio circuits, wireless fidelity (WiFi) modules, power supplies, Bluetooth modules, and other components, which will not be described in detail here.

[0165] In the electronic device shown in FIG11 , the processor 110 may be configured to call a program of a transaction credential review method stored in the memory 120 and specifically perform the following operations:

[0166] Uploading the target model gradient parameters of the local transaction credential audit model to the service platform, wherein the target model gradient parameters are used to instruct the service platform to determine the target model parameters based on the target model gradient parameters uploaded by at least one electronic device;

[0167] receiving the target model parameters sent by the service platform, and updating the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model;

[0168] Transaction credential audit processing is performed based on the target transaction audit model.

[0169] In one embodiment, when executing the step of performing transaction credential review based on the target credential recognition model, the processor 110 specifically performs the following operations:

[0170] Obtain transaction audit credential data of the target user;

[0171] The transaction voucher audit processing is performed on the transaction audit voucher data based on the target transaction voucher audit model to obtain a transaction voucher audit result corresponding to the target user.

[0172] In one embodiment, when the processor 110 performs the step of performing transaction credential audit processing on the transaction audit credential data based on the target transaction credential audit model to obtain the transaction credential audit result corresponding to the target user, the processor 110 specifically performs the following operations:

[0173] Perform voucher classification processing based on the transaction audit voucher data to obtain target transaction audit voucher data, wherein the target transaction audit voucher data carries the target voucher classification corresponding to the transaction audit voucher data;

[0174] The target transaction audit voucher data is subjected to transaction voucher audit processing based on the target transaction voucher audit model to obtain a transaction voucher audit result corresponding to the target user.

[0175] In one embodiment, when the processor 110 performs the step of performing voucher classification processing based on the transaction audit voucher data to obtain target transaction audit voucher data, the processor 110 specifically performs the following operations:

[0176] Perform voucher classification identification processing on the transaction audit voucher data to obtain a target voucher classification corresponding to the transaction audit voucher data;

[0177] The transaction audit voucher data is labeled based on the target voucher classification to obtain target transaction audit voucher data.

[0178] In one embodiment, when the processor 110 performs the step of performing voucher classification processing based on the transaction audit voucher data to obtain target transaction audit voucher data, the processor 110 specifically performs the following operations:

[0179] The transaction audit voucher data is input into a voucher classification and recognition model to obtain target transaction audit voucher data. The voucher classification and recognition model is trained with sample target transaction audit voucher data labels corresponding to labeled sample transaction audit voucher data.

[0180] In one embodiment, before uploading the target model gradient parameters of the local transaction credential audit model to the service platform, the processor 110 further performs the following operations:

[0181] A local transaction credential audit model is determined, and a target model gradient parameter of the local transaction credential audit model is obtained.

[0182] In one embodiment, when executing the step of obtaining the target model gradient parameter of the local transaction credential audit model, the processor 110 specifically performs the following operations:

[0183] Obtain historical transaction audit credential data of historical users;

[0184] The local transaction credential audit model is trained based on the historical transaction audit credential data to obtain target model gradient parameters of the local transaction credential audit model.

[0185] In one embodiment, when the processor 110 performs the step of performing model training processing on the local transaction credential audit model based on the historical transaction audit credential data to obtain the target model gradient parameter of the local transaction credential audit model, the processor 110 specifically performs the following operations:

[0186] If it is determined that the historical transaction audit voucher data meets the preset model training conditions, the local transaction voucher audit model is trained based on the historical transaction audit voucher data to obtain the target model gradient parameters of the local transaction voucher audit model.

[0187] In one embodiment, when executing the step of determining whether the historical transaction audit credential data meets the preset model training condition, the processor 110 specifically performs the following operations:

[0188] Determining whether the accumulated amount of the historical transaction audit voucher data is greater than a preset training amount, and if the accumulated amount is greater than the preset training amount, determining that the historical transaction audit voucher data meets the preset model training condition; or,

[0189] It is determined whether the acquisition time of the historical transaction audit voucher data is within a preset training time period. If the acquisition time is within the preset training time period, it is determined that the historical transaction audit voucher data meets the preset model condition.

[0190] Please refer to Figure 12, which shows a schematic diagram of the structure of a service platform provided by an exemplary embodiment of this specification. The service platform described in this specification may include one or more of the following components: a processor 110, a memory 120, an input device 130, an output device 140, and a bus 150. The processor 110, the memory 120, the input device 130, and the output device 140 may be connected via the bus 150.

[0191] The processor 110 may include one or more processing cores. The processor 110 utilizes various interfaces and circuits to connect various components within the terminal. It executes instructions, programs, code sets, or instruction sets stored in the memory 120, as well as accesses data stored in the memory 120, to perform various functions of the terminal 100 and process data. Optionally, the processor 110 may be implemented using at least one of the following hardware forms: a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). The processor 110 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem handles wireless communications. It is understood that the modem may not be integrated into the processor 110 and may be implemented separately via a communications chip.

[0192] The memory 120 may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory 120 includes a non-transitory computer-readable storage medium. The memory 120 may be used to store instructions, programs, codes, code sets, or instruction sets. The memory 120 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc. The operating system may be an Android system, including a system deeply developed based on the Android system, an iOS system developed by Apple, including a system deeply developed based on the iOS system, or other systems.

[0193] In order for the operating system to distinguish the specific application scenarios of third-party applications, it is necessary to open up data communication between third-party applications and the operating system so that the operating system can obtain the current scenario information of third-party applications at any time, and then perform targeted system resource adaptation based on the current scenario.

[0194] The input device 130 is used to receive input commands or data and includes, but is not limited to, a keyboard, a mouse, a camera, a microphone, or a touch-sensitive device. The output device 140 is used to output commands or data and includes, but is not limited to, a display device and a speaker. In one example, the input device 130 and the output device 140 may be combined, and the input device 130 and the output device 140 may be a touch-sensitive display.

[0195] The touch display screen can be designed as a full screen, a curved screen or a special-shaped screen. The touch display screen can also be designed as a combination of a full screen and a curved screen, or a combination of a special-shaped screen and a curved screen, which is not limited in the embodiments of this specification.

[0196] In addition, those skilled in the art will understand that the structure of the electronic device shown in the above figures does not constitute a limitation on the service platform. The service platform may include more or fewer components than shown, or a combination of certain components, or a different arrangement of components. For example, the electronic device also includes a radio frequency circuit, an input unit, a sensor, an audio circuit, a wireless fidelity (WiFi) module, a power supply, a Bluetooth module, and other components, which will not be detailed here.

[0197] In the service platform shown in FIG11 , the processor 110 may be configured to call a program of a transaction credential review method stored in the memory 120 and specifically perform the following operations:

[0198] receiving target model gradient parameters uploaded by at least one electronic device;

[0199] Determining target model parameters based on the target model gradient parameters uploaded by each of the electronic devices;

[0200] The target model parameters are sent to each of the electronic devices, and the target model parameters are used to instruct each of the electronic devices to update the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model, and perform transaction credential audit processing based on the target transaction credential audit model.

[0201] In one embodiment, when executing the step of determining the target model parameters based on the target model gradient parameters uploaded by each electronic device, the processor 110 specifically performs the following operations:

[0202] Determine a target model gradient mean based on the target model gradient parameters uploaded by each of the electronic devices;

[0203] The local model parameters are updated based on the target model gradient mean to obtain the target model parameters.

[0204] In one embodiment, when executing the step of determining the target model gradient mean based on the target model gradient parameters uploaded by each electronic device, the processor 110 specifically performs the following operations:

[0205] Obtaining the number of samples corresponding to the target model gradient parameter, and determining the target weight corresponding to the target model gradient parameter based on the number of samples;

[0206] A target model gradient mean is calculated based on the target model gradient parameter and the target weight.

[0207] An embodiment of this specification also provides a computer program product, which stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the transaction credential audit method described in the above embodiments.

[0208] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.), and signals involved in the embodiments of this specification are all authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the transaction audit voucher data, target transaction audit voucher data, and historical transaction audit voucher data involved in this specification are all obtained with full authorization.

[0209] Those skilled in the art will appreciate that in one or more of the above examples, the functions described in the embodiments of this specification can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of computer programs from one place to another. The storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0210] The above description is only an optional embodiment of this specification and is not intended to limit this specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of this specification should be included in the scope of protection of this specification.

[0211] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A transaction credential review method, applied to an electronic device, the method comprising: Uploading a target model gradient parameter of a local transaction credential audit model to a service platform, wherein the target model gradient parameter is used to instruct the service platform to determine a target model parameter based on the target model gradient parameter uploaded by at least one of the electronic devices; Receiving the target model parameters sent by the service platform, and updating the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model; Transaction credential audit processing is performed based on the target transaction audit model.

2. According to the method of claim 1, the process of performing transaction voucher auditing based on the target voucher recognition model comprises: Obtain transaction audit credential data of the target user; The transaction credential audit data is processed based on the target transaction credential audit model to obtain a transaction credential audit result corresponding to the target user.

3. The method according to claim 2, wherein the step of performing transaction credential audit processing on the transaction audit credential data based on the target transaction credential audit model to obtain a transaction credential audit result corresponding to the target user comprises: Perform voucher classification processing based on the transaction audit voucher data to obtain target transaction audit voucher data, wherein the target transaction audit voucher data carries the target voucher classification corresponding to the transaction audit voucher data; The target transaction audit voucher data is subjected to transaction voucher audit processing based on the target transaction voucher audit model to obtain a transaction voucher audit result corresponding to the target user.

4. The method according to claim 3, wherein the step of performing voucher classification processing based on the transaction audit voucher data to obtain target transaction audit voucher data comprises: Perform voucher classification identification processing on the transaction audit voucher data to obtain a target voucher classification corresponding to the transaction audit voucher data; The transaction audit voucher data is labeled based on the target voucher classification to obtain target transaction audit voucher data.

5. The method according to claim 3, wherein the step of performing voucher classification processing based on the transaction audit voucher data to obtain target transaction audit voucher data comprises: The transaction audit voucher data is input into a voucher classification and recognition model to obtain target transaction audit voucher data. The voucher classification and recognition model is trained with sample target transaction audit voucher data labels corresponding to labeled sample transaction audit voucher data.

6. The method according to claim 1, before uploading the target model gradient parameters of the local transaction credential audit model to the service platform, further comprises: A local transaction credential audit model is determined, and a target model gradient parameter of the local transaction credential audit model is obtained.

7. According to the method of claim 6, the step of obtaining the target model gradient parameter of the local transaction credential audit model comprises: Obtain historical transaction audit credential data of historical users; Model training processing is performed on the local transaction credential audit model based on the historical transaction audit credential data to obtain target model gradient parameters of the local transaction credential audit model.

8. The method according to claim 7, wherein the step of performing model training processing on the local transaction credential audit model based on the historical transaction audit credential data to obtain a target model gradient parameter of the local transaction credential audit model comprises: If it is determined that the historical transaction audit voucher data meets the preset model training condition, the local transaction voucher audit model is trained based on the historical transaction audit voucher data to obtain the local transaction voucher audit model. Target model gradient parameters for the kernel model.

9. The method according to claim 8, wherein determining that the historical transaction audit voucher data meets a preset model training condition comprises: Determine whether the accumulated amount of the historical transaction audit voucher data is greater than a preset training amount, and if the accumulated amount is greater than the preset training amount, determine that the historical transaction audit voucher data meets the preset model training condition; or, It is determined whether the acquisition time of the historical transaction audit voucher data is within a preset training time period. If the acquisition time is within the preset training time period, it is determined that the historical transaction audit voucher data meets the preset model condition.

10. A transaction voucher review method, applied to a service platform, comprising: Receiving a target model gradient parameter uploaded by at least one electronic device; Determine a target model parameter based on the target model gradient parameter uploaded by each of the electronic devices; The target model parameters are sent to each of the electronic devices, and the target model parameters are used to instruct each of the electronic devices to update a local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model, and perform transaction credential audit processing based on the target transaction credential audit model.

11. The method according to claim 10, wherein determining the target model parameters based on the target model gradient parameters uploaded by each of the electronic devices comprises: Determine a target model gradient mean based on the target model gradient parameters uploaded by each of the electronic devices; The local model parameters are updated based on the target model gradient mean to obtain the target model parameters.

12. The method according to claim 11, wherein determining the target model gradient mean based on the target model gradient parameters uploaded by each of the electronic devices comprises: Obtaining the number of samples corresponding to the target model gradient parameter, and determining the target weight corresponding to the target model gradient parameter based on the number of samples; A target model gradient mean is calculated based on the target model gradient parameter and the target weight.

13. A transaction credential review device, applied to an electronic device, comprising: A parameter uploading module, used to upload the target model gradient parameters of the local transaction credential audit model to the service platform, wherein the target model gradient parameters are used to instruct the service platform to determine the target model parameters based on the target model gradient parameters uploaded by at least one of the electronic devices; A model updating module, configured to receive the target model parameters sent by the service platform, and update the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model; A data processing module is used to perform transaction credential audit processing based on the target transaction audit model.

14. A transaction credential review device, applied to a service platform, comprising: A data receiving module, used to receive target model gradient parameters uploaded by at least one electronic device; A data processing module, used for determining target model parameters based on the target model gradient parameters uploaded by each of the electronic devices; A data sending module is used to send the target model parameters to each of the electronic devices, and the target model parameters are used to instruct each of the electronic devices to update the local transaction credential audit model based on the target model parameters to obtain a target transaction credential audit model, and perform transaction credential audit processing based on the target transaction credential audit model.

15. A computer storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing the method steps of any one of claims 1 to 9 or 10 to 12.

16. A computer program product, wherein the computer program product stores at least one instruction, wherein the at least one instruction is loaded by a processor and executes the method steps according to any one of claims 1 to 9 or 10 to 12.

17. An electronic device comprising: A processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the method steps as claimed in any one of claims 1 to 9 or 10 to 12.

Images