Information classification method and device, equipment and storage medium
Through the neural network model, the privacy and security and transaction activity problems caused by the leakage of official card transaction information is solved, and information security and transaction volume is improved.
Patent Information
- Application Number
- CN202411129559.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-12-17
AI Technical Summary
In the use scenarios of official cards, the leakage of personal consumption information will affect the privacy and security, which will affect the transaction volume and activity of the card, resulting in wasting card resources.
The type of payment transaction is predicted through a neural network model, and classified transaction information based on predefined tags (personal consumption tags and official expenditure tags), and send transaction tags to the deduction server to hide personal consumption records during query and protect privacy information.
It effectively improves the transaction volume and activity of personal official cards, reduces the waste of card resources, and ensures the privacy and security of personal consumption information.
Smart Images
Figure CN119151536A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present application relate to the field of artificial intelligence technology, and in particular to an information classification method, device, equipment and storage medium. Background Art
[0002] With the gradual diversification of payment scenarios, in order to improve payment efficiency in different scenarios, business cards are divided into personal business cards and unit business cards. Among them, personal business cards can be used for official expenses as well as personal consumption.
[0003] In the processing scenario of official expenditure information, such as the reimbursement scenario of personal official cards, the processing terminal can not only query the official expenditure information of the personal official card, but also query the personal consumption information of the personal official card, which leads to the leakage of privacy information such as personal consumption information, thereby affecting the security of privacy information; further, due to the risk of privacy leakage of personal official cards, people who hold personal official cards rarely use their personal official cards for personal consumption in order to protect their personal privacy, resulting in the personal consumption potential of personal official cards not being fully utilized, which in turn causes low transaction volume and activity of personal official cards, resulting in waste of card resources. Summary of the invention
[0004] The embodiments of the present application provide an information classification method, apparatus, device and storage medium to improve the security of privacy information, while increasing the transaction volume and activity of personal business cards, thereby avoiding waste of card resources.
[0005] On the one hand, an embodiment of the present application provides a transaction information classification method, which is applied to a payment platform, and the method includes:
[0006] Obtain transaction information of payment transactions of the target account;
[0007] Predicting a target transaction type corresponding to the payment transaction based on the transaction information by a neural network model;
[0008] Based on the target transaction type, selecting a transaction tag of the payment transaction from a plurality of predefined tags, the plurality of predefined tags comprising: a personal consumption tag and a business expenditure tag;
[0009] The transaction tag of the payment transaction is sent to the deduction server, so that the deduction server responds to the transaction query request for the target account according to the saved transaction tags of each payment transaction of the target account.
[0010] On the one hand, an embodiment of the present application provides a transaction information classification method, which is applied to a deduction server, and the method includes:
[0011] Receiving and saving a transaction tag of a payment transaction sent by a payment platform, wherein the transaction tag of the payment transaction is selected by the payment platform from a plurality of predefined tags based on a target transaction type of the payment transaction, wherein the plurality of predefined tags include: a personal consumption tag and a business expenditure tag; the target transaction type of the payment transaction is predicted by the payment platform through a neural network model based on transaction information of the payment transaction;
[0012] The method further comprises: responding to a transaction query request for the target account according to the stored transaction tags of each payment transaction of the target account.
[0013] On the one hand, an embodiment of the present application provides a transaction information classification device, which is applied to a payment platform, and the device includes:
[0014] The first acquisition module is used to acquire transaction information of a payment transaction of a target account.
[0015] A first processing module, configured to predict a target transaction type corresponding to the payment transaction based on the transaction information by using a neural network model;
[0016] Based on the target transaction type, a transaction tag of the payment transaction is selected from a plurality of predefined tags, wherein the plurality of predefined tags include: a personal consumption tag and a business expenditure tag.
[0017] The first processing module is further used to send the transaction tag of the payment transaction to the deduction server, so that the deduction server responds to the transaction query request for the target account according to the saved transaction tags of each payment transaction of the target account.
[0018] Optionally, the first processing module is further used for:
[0019] After selecting the transaction tag of the payment transaction from a plurality of predefined tags based on the target transaction type, the transaction tag of the payment transaction is returned to the first terminal so that the first terminal displays the transaction tag of the payment transaction.
[0020] Optionally, the neural network model includes: an input layer, an output layer and multiple hidden layers, and the layers are fully connected.
[0021] The first processing module is specifically used for:
[0022] The training data is used to iteratively train the network to be trained until the iteration stop condition is met to obtain the neural network model, wherein each iterative training process includes the following steps:
[0023] Select batch data from the training data and input them into the network to be trained for prediction, so as to obtain corresponding predicted transaction types;
[0024] Determine a loss value based on the obtained predicted transaction type, the actual transaction type corresponding to the batch data, and the loss function;
[0025] With the goal of minimizing the loss value, the parameter information of the network to be trained is adjusted.
[0026] On the one hand, an embodiment of the present application provides a transaction information classification device, which is applied to a deduction server, and the device includes:
[0027] The second acquisition module is used to receive and save the transaction tag of the payment transaction sent by the payment platform, the transaction tag of the payment transaction is selected by the payment platform from multiple predefined tags based on the target transaction type of the payment transaction, and the multiple predefined tags include: personal consumption tags and official expenditure tags; the target transaction type of the payment transaction is predicted by the payment platform through a neural network model based on the transaction information of the payment transaction.
[0028] The second processing module is used to respond to the transaction query request for the target account according to the saved transaction tags of each payment transaction of the target account.
[0029] Optionally, the second processing module is specifically configured to:
[0030] Receiving a transaction query request for corporate expenditures of the target account sent by the second terminal;
[0031] Acquire at least one payment transaction corresponding to the official expenditure tag from the multiple payment transactions of the target account;
[0032] Returning a transaction record of the at least one payment transaction to the second terminal.
[0033] Optionally, the transaction query request carries a transaction time period;
[0034] The second processing module is specifically used for:
[0035] At least one payment transaction that is within the transaction time period and corresponds to the official expenditure tag is obtained from the multiple payment transactions of the target account.
[0036] Optionally, the second processing module is further used for:
[0037] In response to a modification request for the transaction tag of the payment transaction sent by the first terminal or the payment platform, modify the saved transaction tag of the payment transaction.
[0038] In one aspect, an embodiment of the present application provides a computer device, including:
[0039] A memory for storing program instructions;
[0040] The processor is used to call the program instructions stored in the memory and execute the steps of the transaction information classification method according to the obtained program.
[0041] On the one hand, an embodiment of the present application provides a computer-readable storage medium, including computer-readable instructions. When a computer reads and executes the computer-readable instructions, the computer executes the steps of the above-mentioned transaction information classification method.
[0042] In an embodiment of the present application, the payment platform identifies the transaction tags of the payment transactions based on the transaction information of the payment transactions of the target account through the trained neural network model. The transaction tags of the payment transactions are personal consumption tags and official expenditure tags. The transaction tags of the payment transactions are sent to the deduction server for storage. In this way, the deduction server can respond to the transaction query request of the payment transaction according to the transaction tags of different payment transactions. Therefore, when the deduction server receives the query of the public consumption record of the target account, the personal consumption record of the target account will be hidden, thereby avoiding the risk of personal consumption privacy leakage. Furthermore, since personal privacy is effectively protected, the overall transaction volume and activity of the official card will also increase accordingly, thereby reducing the waste of card resources. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0044] Figure 1 A schematic diagram of a system architecture provided in an embodiment of the present application;
[0045] Figure 2 A flowchart of a transaction information classification method provided in an embodiment of the present application;
[0046] Figure 3 A flowchart of a transaction record query method provided in an embodiment of the present application;
[0047] Figure 4 A flowchart of a transaction information classification method provided in an embodiment of the present application;
[0048] Figure 5 A schematic diagram of the structure of a transaction information classification device provided in an embodiment of the present application;
[0049] Figure 6A schematic diagram of the structure of a transaction information classification device provided in an embodiment of the present application;
[0050] Figure 7 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0051] In order to make the purpose, technical solution and beneficial effects of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0052] For ease of understanding, the terms involved in the embodiments of the present invention are explained below.
[0053] The following is a brief introduction to the system architecture diagram used in the technical solution of the embodiment of the present application. It should be noted that the system architecture diagram introduced below is only used to illustrate the embodiment of the present application and is not a limitation.
[0054] refer to Figure 1 , which is a system architecture diagram applicable to the embodiment of the present application, and the system architecture at least includes a payment platform 101, a first terminal 102, a second terminal 103, and a deduction server 104. The first terminal 102 pre-installs an application with a payment function, in which a personal business card can be bound, and the personal business card supports business expenses and personal consumption. The application can be a client application, a web application, a small program application, etc. The first terminal 102 can be a smart phone, a smart watch, a tablet computer, a laptop computer, etc., but is not limited thereto.
[0055] The payment platform 101 is a backend server of an application with payment functions. The payment platform 101 can provide functions such as transaction settlement, information exchange, settlement data processing, technical support, risk control, and data analysis.
[0056] The second terminal 103 is a pre-installed application for processing official expenditure information, such as personal official card reimbursement. The second terminal 103 can be a smart phone, a smart watch, a tablet computer, a laptop computer, etc., but is not limited thereto.
[0057] The deduction server 104 is a business card support system provided by the issuing bank corresponding to the personal business card, and can provide services such as cloud storage, account expenditure and repayment, and information transmission services.
[0058] In the embodiment of the present application, the payment platform 101 and the first terminal 102, the payment platform 101 and the deduction server 104, the first terminal 102 and the deduction server 104, and the second terminal 103 and the deduction server 104 can be directly or indirectly connected through a network. The network can be a wired network; it can also be a wireless network, for example, the wireless network can be a mobile cellular network, which is not limited in the embodiment of the present application.
[0059] The following is based on Figure 1 The system architecture shown in the figure, the present application embodiment provides a flow chart of a transaction information classification method, such as Figure 2 As shown, the process of the method is interactively executed by the payment platform and the deduction server, and includes the following steps:
[0060] Step 201: The payment platform obtains transaction information of the payment transaction of the target account.
[0061] Specifically, the target account can be a personal business card account bound to the payment platform, or it can be another type of account. During the payment transaction, QR code payment, online payment, etc. are performed through the personal business card account. The transaction information of the payment transaction includes: transaction amount, transaction type, merchant type and other information.
[0062] In step 202, the payment platform predicts the target transaction type corresponding to the payment transaction based on the transaction information through a neural network model.
[0063] Through the neural network model, the target transaction type corresponding to the payment transaction is predicted from the preset multiple transaction types based on the transaction information, and the target transaction type is one of the preset multiple transaction types.
[0064] Specifically, the preset transaction types include: personal consumption expenditure, other corporate expenditure, office expenses, printing fees, consulting fees, handling fees, water and electricity fees, postage and telecommunications fees, property management fees, travel expenses, maintenance (care) fees, rental fees, conference fees, training fees, official reception fees and other transaction types.
[0065] In some embodiments, the neural network model can be a deep neural network (DNN) model, a convolutional neural network model, a recurrent neural network model, etc., which is not limited in the embodiments of the present application.
[0066] The neural network model includes: an input layer, an output layer and multiple hidden layers, and the layers are fully connected. In this application, the neural network model is trained in the following way:
[0067] The neural network model to be trained is iteratively trained using the training data until the iteration stop condition is met to obtain the neural network model, wherein each iterative training process includes the following steps:
[0068] When batch data is selected from the training data and input into the training network for prediction, the corresponding predicted transaction type is obtained; the loss value is determined based on the obtained predicted transaction type, the real transaction type corresponding to the batch data and the loss function; and the parameter information of the neural network to be trained is adjusted with the goal of minimizing the loss value.
[0069] Specifically, the above-mentioned preset multiple transaction types are used as labels to annotate the data set; 90% of the data are randomly selected from the data set as training data, 5% of the data are used as verification data, and 5% of the data are used as test data.
[0070] In each iterative training process, batch data is selected from the training data with or without replacement, and the selected batch data is input into the neural network to be trained (i.e., the neural network model to be trained), and forward calculation is performed layer by layer through the sigmoid activation function until the output layer is passed to obtain the predicted transaction type. The predicted transaction type is compared with the actual transaction type, and the loss value is calculated using the cross entropy loss function, which is used to measure the difference between the predicted value and the actual value.
[0071] In addition, after the loss value is calculated by forward propagation, back propagation is performed according to the chain rule, starting from the output layer, and traversing backwards layer by layer to calculate the weight gradient of the loss function for each layer until the input layer, and obtaining the weight gradient. According to the weight gradient obtained in the reverse process, gradient descent or other optimization strategies are used to optimize the parameters of the neural network model through continuous iteration to minimize the loss value, thereby improving the classification prediction performance of the neural network model.
[0072] After the training is completed, the classification model evaluation method is used to input the verification data into the neural network model for verification. The effect of the neural network model is judged by the four evaluation indicators of accuracy, precision, recall rate and F-1 score. The calculation formulas of the four evaluation indicators are shown in the following formulas (1) to (4):
[0073]
[0074] Among them, Accracy represents accuracy, Precision represents precision, Recall represents recall, and F-1score represents F-1 score; TP represents true positive examples, that is, the actual positive example is predicted to be positive; TN represents true negative examples, that is, the actual negative example is predicted to be negative; FP represents false positive examples, that is, the actual negative example is predicted to be positive; FN represents false negative examples, that is, the actual positive example is predicted to be negative.
[0075] In addition, cross-validation technology is used on the training data and validation data to evaluate the accuracy and stability of the model. 5% of the data other than the 5% test data is randomly divided into validation data and 90% is divided into test data. The accuracy of the validation data is calculated multiple times to evaluate the accuracy and stability of the model, so that the accuracy reaches more than 80%. This method can reduce overfitting to a certain extent and is more conducive to finding parameters suitable for the neural network model. Finally, the test data is input into the neural network model to further verify the accuracy of the model. By adjusting parameters such as learning rate, batch size, and number of iterations (epoch), the final accuracy of the test data reaches 80%. The model with an accuracy of more than 80% after verification and testing is deployed to the payment platform for application.
[0076] In the embodiment of the present application, the use of a neural network model can increase the training speed of the model and at the same time improve the accuracy of transaction information classification.
[0077] Step 203: The payment platform selects a transaction tag for the payment transaction from a plurality of predefined tags based on the target transaction type. The plurality of predefined tags include a personal consumption tag and a business expenditure tag.
[0078] Specifically, when the target transaction type is personal consumption expenditure, the transaction tag of the payment transaction is defined as a personal consumption tag; when the target transaction type is other official expenses or any of the multiple transaction types such as office expenses, printing expenses, consulting fees, etc., the transaction tag of the payment transaction is defined as an official expense tag.
[0079] In an embodiment of the present application, based on the target transaction type, a transaction tag for the payment transaction is selected from a plurality of predefined tags, thereby simplifying the complexity of the transaction tag and improving the user experience.
[0080] In some embodiments, after selecting the transaction tag of the payment transaction from a plurality of predefined tags based on the target transaction type, the payment platform returns the transaction tag of the payment transaction to the first terminal so that the first terminal displays the transaction tag of the payment transaction.
[0081] Specifically, during the payment transaction, the payment platform may return the transaction tag of the payment transaction to the first terminal, and the first terminal may display the transaction tag of the payment transaction in the settlement interface; or after the payment transaction is completed, the payment platform may return the transaction tag of the payment transaction to the first terminal through notification push or other means, and the first terminal may display the transaction tag definition result. The above process can remind the user to know the transaction tag definition result, and that the transaction tag will affect the subsequent query operation of the business card consumption information, and also remind the user that if he has any objection to the transaction tag of the payment transaction, he can submit a transaction tag modification request.
[0082] In some embodiments, the first terminal sends a modification request for the transaction tag of the payment transaction to the payment platform, and the payment platform sends the modification request to the deduction server; of course, the first terminal can also directly send the modification request for the transaction tag of the payment transaction to the deduction server, and the deduction server modifies the saved transaction tag of the payment transaction.
[0083] Specifically, the user can trigger a modification operation of the transaction tag of the payment transaction in the display interface of the first terminal, and the modification operation includes: the modified transaction tag. In response to the modification operation, the first terminal sends a modification request to the deduction server through the payment platform or directly, and after the deduction server agrees, the deduction server adjusts the saved transaction tag of the payment transaction to the modified transaction tag specified by the user.
[0084] In the embodiment of the present application, the payment platform returns the transaction tag of the payment transaction to the first terminal for display, so the user can promptly know the transaction tag classification result of the payment transaction, thereby improving the user experience. In addition, when the user needs to modify the transaction tag, the modification request can be sent to the deduction server through the payment platform or directly, so that the deduction server can modify the saved transaction tag of the payment transaction, thereby improving the accuracy of the transaction tag.
[0085] Step 204: The payment platform sends the transaction tag of the payment transaction to the deduction server.
[0086] Step 205: The deduction server responds to the transaction query request for the target account according to the saved transaction tags of each payment transaction of the target account.
[0087] In an embodiment of the present application, the payment platform identifies the transaction tag of the payment transaction based on the transaction information of the payment transaction of the target account through a trained neural network model. The transaction tags of the payment transaction are personal consumption tags and official expenditure tags. The transaction tags of the payment transaction are sent to the deduction server for storage. In this way, the deduction server can respond to the transaction query request of the payment transaction according to the transaction tags of different payment transactions. Therefore, when the deduction server receives the query of the public consumption record of the target account, the personal consumption record of the target account will be hidden, thereby avoiding the risk of personal consumption privacy leakage. Furthermore, since personal privacy is effectively protected, the overall transaction volume and activity of the official card will also increase accordingly, thereby reducing the waste of card resources.
[0088] In some embodiments, the deduction server responds to the transaction query request process for the target account according to the transaction tags of each payment transaction of the target account stored. Figure 3 As shown, the following steps are included:
[0089] Step 301: The second terminal sends a transaction query request for corporate expenditures of a target account to a deduction server.
[0090] Specifically, the second terminal accesses the deduction server (ie, the business card support system provided by the issuing bank) through the domain name, and inserts the corresponding authority certificate during the access process.
[0091] In addition, the information carried in the transaction query request includes: the name and card number of the cardholder associated with the target account.
[0092] Step 302: The deduction server obtains at least one payment transaction corresponding to the official expenditure tag from multiple payment transactions of the target account.
[0093] Specifically, the deduction server obtains the transaction record of at least one payment transaction corresponding to the official expenditure tag from multiple payment transactions of the target account according to the information carried in the transaction query request, while the payment transaction corresponding to the personal consumption tag is directly hidden.
[0094] Step 303: The deduction server returns a transaction record of at least one payment transaction.
[0095] Specifically, after receiving the transaction record of at least one payment transaction sent by the deduction server, the second terminal can verify the transaction records of payment transactions with multiple official expenditure tags at one time, and form a repayment summary table with the verified transaction records. The second terminal submits the repayment summary table to the deduction server provided by the specific official card issuing bank, and the deduction server offsets or deducts the consumption records of the official expenditure tags of the target account in the repayment summary table.
[0096] In an embodiment of the present application, when the deduction server obtains payment transactions in response to the transaction query request sent by the second terminal, it can only obtain at least one payment transaction corresponding to the official expenditure tag of the target account, while the payment transactions with personal consumption tags in the target account are hidden. In this way, the second terminal's query authority for payment transactions with personal consumption tags is limited, thereby protecting the user's personal consumption privacy, helping to increase the overall transaction volume and activity of official cards, and effectively avoiding waste of card resources.
[0097] In some embodiments, the second terminal sends a transaction query request for the official expenditure of the target account carrying a transaction time period to the deduction server, and the deduction server obtains at least one payment transaction within the transaction time period and corresponding to the official expenditure tag from multiple payment transactions of the target account.
[0098] Specifically, the information carried in the transaction query request also includes: transaction time period. The deduction server batch queries the transaction records of at least one payment transaction corresponding to the official expenditure tag within the transaction time period from multiple payment transactions of the target account. Multiple public transactions can be verified at one time, and official card expenditures can also be monitored.
[0099] In an embodiment of the present application, the second terminal sends a transaction query request for official expenditures of a target account carrying a time period, which can batch query the transaction records of the official expenditure tags of the target account, thereby improving the transaction information query rate and the processing efficiency of financial reimbursement.
[0100] In order to better explain the embodiment of the present application, the following describes a method for classifying transaction information provided by the embodiment of the present application in combination with the business card reimbursement scenario. The process of the method can be as follows: Figure 1 The payment platform, the first terminal, the second terminal and the deduction server are interactively executed, including the following steps: Figure 4 As shown:
[0101] Step 401: The payment platform trains the neural network model.
[0102] The neural network model has strong data fitting capabilities and has advantages in processing nonlinear relationships of similar transaction information, corporate consumption and personal consumption transaction types.
[0103] Step 402: The first terminal binds a personal business card to perform a payment transaction.
[0104] The first terminal is pre-installed with an application with a payment function, which can be bound to a personal business card for payment transactions. The neural network model has an advantage in accurately fitting the nonlinear classification relationship of transaction information.
[0105] Step 403: The payment platform obtains payment transaction information.
[0106] As the backend server of applications with payment functions, the payment platform can obtain payment transaction information of each payment transaction of personal business cards.
[0107] Step 404: The payment platform predicts the transaction type of the payment transaction based on the payment transaction information of the payment transaction through a neural network, and defines a transaction tag of the payment transaction according to the transaction type.
[0108] Step 405: The payment platform sends a transaction tag of the payment transaction to the first terminal.
[0109] Step 406: The payment platform sends a transaction tag of the payment transaction to the deduction server.
[0110] The deduction server is associated with the issuing bank of the personal business card.
[0111] Step 407: The deduction server associates the personal business card with the transaction tag of the payment transaction and saves it.
[0112] Step 408: The first terminal sends a transaction tag modification request to the deduction server.
[0113] Step 409: The deduction server modifies the transaction tag of the saved payment transaction.
[0114] Step 410: The second terminal obtains the reimbursement application.
[0115] Specifically, the reimbursement application includes: the name of the cardholder of the personal business card, card number, transaction amount, time and other information, where the time can be a specific transaction date or a time period restricted to a certain time range.
[0116] Step 411: The second terminal sends a transaction query request for the personal business card to the deduction server.
[0117] The transaction inquiry request includes: the cardholder name and card number in the reimbursement application.
[0118] Step 412: The deduction server obtains at least one payment transaction corresponding to the official expenditure tag from multiple payment transactions of the personal official card.
[0119] Specifically, when the reimbursement application does not provide reimbursement information such as specific date and amount, the second terminal has the right to send a batch query request for the payment transaction information of all official expense tags of the personal official card to the deduction server, thereby realizing real-time monitoring of the payment information of the official expense tags of the personal official card, and improving the transaction risk and compliance control level of the official card.
[0120] In addition, the transaction query request sent by the second terminal to the deduction server based on the reimbursement application may not specify a specific date, but carry a transaction time period in the transaction query request. The deduction server obtains at least one payment transaction corresponding to the official expenditure tag based on the cardholder name, card number and transaction time period information associated with the personal official card provided in the transaction query request.
[0121] Step 413: The deduction server returns a transaction record of at least one payment transaction to the second terminal.
[0122] Step 414: The second terminal generates a repayment summary table based on the obtained transaction records.
[0123] Step 415: The second terminal submits the repayment summary table to the deduction server to complete the repayment.
[0124] The transaction query request sent by the second terminal to the deduction server carries a time period, so that the deduction server can batch query the transaction records of payment transactions of personal business card official expenditure tags within a certain time range, thereby improving the transaction information query efficiency of the deduction server.
[0125] In the embodiment of the present application, the payment platform uses the neural network model to calculate the target transaction type corresponding to the payment transaction based on the transaction information of the payment transaction, and obtains the transaction tag of the payment transaction based on the transaction type. In this way, the transaction type of each payment transaction is accurately distinguished, and the deduction server can respond differently according to the transaction tag when receiving the transaction query request, so that the payment transaction with the personal consumption tag is not leaked, which effectively protects the privacy of personal consumption, and the transaction volume and activity of the official card are also increased accordingly, thereby reducing the waste of card resources. Through the transaction query request carrying the time period sent by the second terminal to the deduction server, the transaction records of payment transactions with multiple official expenditure tags can be verified at one time, which improves the efficiency of financial reimbursement and can also complete the real-time monitoring of the public consumption records of personal official cards.
[0126] Based on the same technical concept, the present application embodiment provides a structural diagram of a transaction information classification device, which is applied to a payment platform, such as Figure 5 As shown, the transaction information classification device 500 includes:
[0127] The first acquisition module 501 is used to acquire transaction information of a payment transaction of a target account.
[0128] A first processing module 502, configured to predict a target transaction type corresponding to the payment transaction based on the transaction information by using a neural network model;
[0129] Based on the target transaction type, a transaction tag of the payment transaction is selected from a plurality of predefined tags, wherein the plurality of predefined tags include: a personal consumption tag and a business expenditure tag.
[0130] The first processing module 502 is further configured to send the transaction tag of the payment transaction to the deduction server, so that the deduction server responds to the transaction query request for the target account according to the stored transaction tags of each payment transaction of the target account.
[0131] Optionally, the first processing and sending module 502 is further used for:
[0132] After selecting the transaction tag of the payment transaction from a plurality of predefined tags based on the target transaction type, the transaction tag of the payment transaction is returned to the first terminal so that the first terminal displays the transaction tag of the payment transaction.
[0133] Optionally, the neural network model includes: an input layer, an output layer and multiple hidden layers, and the layers are fully connected.
[0134] The first processing module 502 is specifically used for:
[0135] The training data is used to iteratively train the network to be trained until the iteration stop condition is met to obtain the neural network model, wherein each iterative training process includes the following steps:
[0136] Select batch data from the training data and input them into the network to be trained for prediction, so as to obtain corresponding predicted transaction types;
[0137] Determine a loss value based on the obtained predicted transaction type, the actual transaction type corresponding to the batch data, and the loss function;
[0138] With the goal of minimizing the loss value, the parameter information of the network to be trained is adjusted.
[0139] Based on the same technical concept, the embodiment of the present application provides a structural diagram of a transaction information classification device, which is applied to a deduction server, such as Figure 6 As shown, the transaction information classification device 600 includes:
[0140] The second acquisition module 601 is used to receive and save the transaction tag of the payment transaction sent by the payment platform, the transaction tag of the payment transaction is selected by the payment platform from multiple predefined tags based on the target transaction type of the payment transaction, and the multiple predefined tags include: personal consumption tags and official expenditure tags; the target transaction type of the payment transaction is predicted by the payment platform through a neural network model based on the transaction information of the payment transaction.
[0141] The second processing module 602 is used to respond to the transaction query request for the target account according to the saved transaction tags of each payment transaction of the target account.
[0142] Optionally, the second processing module 602 is specifically configured to:
[0143] Receiving a transaction query request for corporate expenditures of the target account sent by the second terminal;
[0144] Acquire at least one payment transaction corresponding to the official expenditure tag from the multiple payment transactions of the target account;
[0145] Returning a transaction record of the at least one payment transaction to the second terminal.
[0146] Optionally, the transaction query request carries a transaction time period;
[0147] The second processing module 602 is specifically used for:
[0148] At least one payment transaction that is within the transaction time period and corresponds to the official expenditure tag is obtained from the multiple payment transactions of the target account.
[0149] Optionally, the second processing module 602 is further configured to:
[0150] In response to a modification request for the transaction tag of the payment transaction sent by the first terminal or the payment platform, modify the saved transaction tag of the payment transaction.
[0151] In an embodiment of the present application, the neural network model deployed in the payment platform predicts the transaction information of the payment transaction and generates a transaction tag result, and can accurately distinguish the payment information of the official card payment transaction with the official expenditure tag and the personal consumption tag, so that the deduction server can only obtain the transaction information of the official expenditure tag of the target account in response to the query request of the second terminal for the official expenditure tag of the target account, while the transaction information of the personal consumption tag is hidden, which effectively protects the user's personal consumption privacy, thereby promoting the user's willingness to use personal official cards for personal consumption, which is conducive to improving the overall transaction volume and activity of personal official cards; real-time monitoring of personal official cards' public consumption improves the risk and regulatory control level of the deduction server in managing personal official card transaction information.
[0152] Based on the same technical concept, the embodiment of the present application provides a computer device, which can be Figure 1 The terminal device and / or server shown, such as Figure 7 As shown, it includes at least one processor 701 and a memory 702 connected to the at least one processor. The specific link medium between the processor 701 and the memory 702 is not limited in the embodiment of the present application. Figure 7 For example, the processor 701 and the memory 703 are connected via a bus. The bus can be divided into an address bus, a data bus, a control bus, etc.
[0153] In the embodiment of the present application, the memory 702 stores instructions executed by at least one processor 701, and the at least one processor 701 can perform the steps of the above-mentioned transaction information classification method by executing the instructions stored in the memory 702.
[0154] The processor 701 is the control center of the computer device, and can use various interfaces and lines to connect various parts of the computer device, and realize the classification of transaction information by running or executing instructions stored in the memory 702 and calling data stored in the memory 702. Optionally, the processor 701 may include one or more processing modules, and the processor 701 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communication. It is understandable that the above-mentioned modem processor may not be integrated into the processor 701. In some embodiments, the processor 701 and the memory 702 may be implemented on the same chip, and in some embodiments, they may also be implemented separately on independent chips.
[0155] Processor 701 can be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application-specific integrated circuit (ASIC), a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, and can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor can be a microprocessor or any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware processor for execution, or can be executed by a combination of hardware and software modules in the processor.
[0156] The memory 702 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules. The memory 702 may include at least one type of storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (Random Access Memory, RAM), a static random access memory (Static Random Access Memory, SRAM), a programmable read-only memory (Programmable Read Only Memory, PROM), a read-only memory (Read Only Memory, ROM), an electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), a magnetic memory, a disk, an optical disk, etc. The memory 702 is any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer device, but is not limited thereto. The memory 702 in the embodiment of the present application can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.
[0157] Based on the same inventive concept, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program that can be executed by a computer device. When the program runs on the computer device, the computer device executes the steps of the above-mentioned transaction information classification method.
[0158] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0159] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0160] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0161] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0162] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.
Claims
1. A transaction information classification method, characterized in that: Applied to payment platforms, including: Obtain transaction information of payment transactions of the target account; Predicting a target transaction type corresponding to the payment transaction based on the transaction information by a neural network model; Based on the target transaction type, selecting a transaction tag of the payment transaction from a plurality of predefined tags, the plurality of predefined tags comprising: a personal consumption tag and a business expenditure tag; The transaction tag of the payment transaction is sent to the deduction server, so that the deduction server responds to the transaction query request for the target account according to the saved transaction tags of each payment transaction of the target account.
2. The method according to claim 1, characterized in that After selecting the transaction tag of the payment transaction from a plurality of predefined tags based on the target transaction type, the method further includes: The payment platform returns the transaction tag of the payment transaction to the first terminal, so that the first terminal displays the transaction tag of the payment transaction.
3. The method according to claim 2, characterized in that Also includes: In response to the modification request for the transaction tag of the payment transaction sent by the first terminal, the modification request is sent to the deduction server, so that the deduction server modifies the saved transaction tag of the payment transaction.
4. The method according to claim 1, characterized in that The neural network model is trained in the following way: The training data is used to iteratively train the network to be trained until the iteration stop condition is met to obtain the neural network model, wherein each iterative training process includes the following steps: Select batch data from the training data and input them into the network to be trained for prediction, so as to obtain corresponding predicted transaction types; Determine a loss value based on the obtained predicted transaction type, the actual transaction type corresponding to the batch data, and the loss function; With the goal of minimizing the loss value, the parameter information of the network to be trained is adjusted.
5. The method according to claim 4, characterized in that The neural network model includes: an input layer, an output layer and multiple hidden layers, and the layers are fully connected.
6. A transaction information classification method, characterized in that: Applied to the deduction server, including: Receiving and saving a transaction tag of a payment transaction sent by a payment platform, wherein the transaction tag of the payment transaction is selected by the payment platform from a plurality of predefined tags based on a target transaction type of the payment transaction, wherein the plurality of predefined tags include: a personal consumption tag and a business expenditure tag; the target transaction type of the payment transaction is predicted by the payment platform through a neural network model based on transaction information of the payment transaction; The method further comprises: responding to a transaction query request for the target account according to the stored transaction tags of each payment transaction of the target account.
7. The method according to claim 6, characterized in that The step of responding to a transaction query request for the target account according to the saved transaction tags of each payment transaction of the target account includes: Receiving a transaction query request for corporate expenditures of the target account sent by the second terminal; Acquire at least one payment transaction corresponding to the official expenditure tag from the multiple payment transactions of the target account; Returning a transaction record of the at least one payment transaction to the second terminal.
8. The method according to claim 7, characterized in that The transaction query request carries a transaction time period; The acquiring at least one payment transaction corresponding to the official expenditure tag from the multiple payment transactions of the target account includes: At least one payment transaction that is within the transaction time period and corresponds to the official expenditure tag is obtained from the multiple payment transactions of the target account.
9. The method according to claim 6, characterized in that Also includes: In response to a modification request for the transaction tag of the payment transaction sent by the first terminal or the payment platform, modify the saved transaction tag of the payment transaction.
10. A device for classifying transaction information, characterized in that: Applied to payment platforms, including: A first acquisition module, used to acquire transaction information of a payment transaction of a target account; A first processing module, configured to predict a target transaction type corresponding to the payment transaction based on the transaction information by using a neural network model; Based on the target transaction type, selecting a target transaction tag of the payment transaction from a plurality of predefined tags, the plurality of predefined tags comprising: a personal consumption tag and a business expenditure tag; The first processing module is also used to send the target transaction tag of the payment transaction to the deduction server, so that the deduction server responds to the query request for the official expenditure transaction of the target account according to the target transaction tags of each payment transaction saved in the target account.
11. A device for classifying transaction information, characterized in that: Applied to the deduction server, including: A second acquisition module is used to receive and save a transaction tag of a payment transaction sent by a payment platform, wherein the transaction tag of the payment transaction is selected by the payment platform from a plurality of predefined tags based on a target transaction type of the payment transaction, wherein the plurality of predefined tags include: a personal consumption tag and a business expenditure tag; the target transaction type of the payment transaction is obtained by the payment platform through prediction based on the transaction information of the payment transaction through a neural network model; The second processing module is used to respond to the transaction query request for the target account according to the saved transaction tags of each payment transaction of the target account.
12. A computer device, characterized in that: include: A memory for storing program instructions; A processor, configured to call the program instructions stored in the memory, and execute the method according to any one of claims 1 to 9 according to the obtained program.
13. A computer-readable storage medium, characterized in that: The method comprises computer-readable instructions, and when a computer reads and executes the computer-readable instructions, the computer is caused to execute the method according to any one of claims 1 to 9.