Data processing method and device, storage medium and electronic equipment

Abstract

The invention discloses a data processing method and device, a storage medium and electronic equipment, and the method comprises the steps: obtaining first graph structure data corresponding to a target transaction, determining at least one transaction sub-graph structure data corresponding to the first graph structure data, determining a first transaction graph processing model based on the transaction sub-graph structure data, performing edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model, obtaining second graph structure data of the target transaction, and performing graph sparsification processing on the second graph structure data based on the second transaction graph processing model to obtain target transaction graph structure data, and performing target transaction processing based on the target transaction graph structure data.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a data processing method, device, storage medium and electronic device. Background Art

[0002] In practical applications, graph data is widely used in various business fields. Graph data draws entities and the relationships between entities into a complex network diagram (also called graph data). For example, the graph data corresponding to financial trading firms will record and analyze thousands of transactions in the financial market, while the graph data corresponding to social networks connects the complex personal relationships between hundreds of millions of users. The graph data under such financial trading networks and social networks often contains more than 100 million nodes and a huge number of tens of billions of edges, forming a highly complex network structure. Graph data may contain important information such as user behavior preferences, social influence, and financial risk transmission. Summary of the invention

[0003] This specification provides a data processing method, device, storage medium and electronic device, and the technical solution is as follows:

[0004] In a first aspect, this specification provides a data processing method, the method comprising:

[0005] Acquire first graph structure data corresponding to the target transaction, and determine at least one transaction subgraph structure data corresponding to the first graph structure data;

[0006] Determine a first transaction graph processing model based on the transaction subgraph structure data, perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model, wherein the second transaction graph processing model is a lightweight model relative to the first transaction graph processing model;

[0007] Acquire the second graph structure data of the target transaction, and perform graph sparse processing on the second graph structure data based on the second transaction graph processing model to obtain the target transaction graph structure data;

[0008] Target transaction processing is performed based on the target transaction graph structure data.

[0009] In a second aspect, this specification provides a data processing device, the device comprising:

[0010] A data acquisition module, used to acquire first graph structure data corresponding to a target transaction, and determine at least one transaction subgraph structure data corresponding to the first graph structure data;

[0011] a knowledge distillation module, configured to determine a first transaction graph processing model based on the transaction subgraph structure data, and perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model, wherein the second transaction graph processing model is a lightweight model relative to the first transaction graph processing model;

[0012] A graph sparsification module, used for acquiring the second graph structure data of the target transaction, and performing graph sparsification processing on the second graph structure data based on the second transaction graph processing model to obtain the target transaction graph structure data;

[0013] A transaction processing module is used to perform target transaction processing based on the target transaction graph structure data.

[0014] In a third aspect, the present specification provides a computer storage medium, wherein the computer storage medium stores at least one instruction, wherein the instruction is suitable for being loaded by a processor and executing the method steps of one or more embodiments of the present specification.

[0015] In a fourth aspect, the present specification provides a computer program product, wherein the computer program product stores at least one instruction, wherein the instruction is suitable for being loaded by a processor and executing the method steps of one or more embodiments of the present specification.

[0016] In a fifth aspect, the present specification provides an electronic device, which may include: 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 of one or more embodiments of the present specification.

[0017] The beneficial effects brought by the technical solutions provided by some embodiments of this specification include at least:

[0018] In one or more embodiments of the present specification, first graph structure data corresponding to the target transaction is obtained, transaction subgraph structure data corresponding to the first graph structure data is determined, a first transaction graph processing model is determined based on the transaction subgraph structure data, edge structure knowledge distillation is performed on the first transaction graph processing model to obtain a second transaction graph processing model, and graph sparsification is performed on the second graph structure data of the target transaction based on the second transaction graph processing model to obtain target transaction graph structure data. By combining the expressiveness of the first graph structure data of the target transaction through deep learning and the computational efficiency of the lightweight second transaction graph processing model, the contradiction between computing performance and model accuracy of large-scale graph sparsification is resolved. Subsequently, target transaction processing can be performed based on the target transaction graph structure data, showing significant effects in improving model performance, increasing reasoning speed, and reducing computing resource consumption, which helps to promote technological innovation and commercial applications in related transaction scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in this specification or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 It is a scenario diagram of a data processing system provided in this specification;

[0021] Figure 2 It is a flowchart of a data processing method provided in this specification;

[0022] Figure 3 It is a schematic diagram of a scenario of graph structure data processing provided in this specification;

[0023] Figure 4 This is a flow chart of an edge structure knowledge distillation method provided in this specification;

[0024] Figure 5 It is a flowchart of a first transaction graph processing model provided in this specification;

[0025] Figure 6 It is a flowchart of a second transaction graph processing model provided in this specification;

[0026] Figure 7 It is a flowchart of a graph sparse processing provided in this specification;

[0027] Figure 8 It is a structural schematic diagram of a data processing device provided in this specification;

[0028] Fig. 9 It is a structural schematic diagram of an electronic device provided in this manual. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in this specification to clearly and completely describe the technical solutions in this specification. Obviously, the described embodiments 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 ordinary technicians in this field without creative work are within the scope of protection of this specification.

[0030] In the description of this specification, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In the description of this specification, it should be noted that, unless otherwise clearly specified and limited, "including" and "having" and any of their variations 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 optionally also includes steps or units that are not listed, or optionally also includes 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 in 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 there can be three relationships, for example, A and / or B, which 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 an "or" relationship.

[0031] In related technologies, learning and reasoning on graph-structured data is very difficult, partly due to its large scale and complex structure. At the same time, the requirements for relationship modeling in different transaction scenarios vary greatly, and different applications often need to optimize the base map construction method of their graph-structured data for specific tasks. For example, in the recommendation system task, the interactive relationship between users may be a valuable connection, while in the task of identifying fake account behavior, the same connections may appear insignificant or even interference factors. In addition, as time changes, the network base map data of the graph-structured data will be dynamically updated in real time, outdated edge relationships may disappear, and new relationships are constantly generated. This daily dynamic change requires the ability to respond quickly and maintain and update the graph-structured data in a rapidly changing situation.

[0032] When processing these graph-structured data (which can be called large-scale graphs), the sparsification of graph-structured data becomes particularly critical. Graph sparsification can be understood as the process of reducing non-core edges while retaining the core characteristics of the graph data structure. This can not only greatly improve the computational efficiency of subsequent tasks, but also enhance the transaction model (such as content recommendation model, credit processing model) to grasp key information, thereby to a certain extent reducing the computational and storage burden brought by processing large-scale graphs. However, traditional graph sparsification methods often rely on heuristic rules or predefined thresholds. Such processing methods are not only difficult to adapt to the personalized needs of different transaction scenarios, but also difficult to flexibly adapt to the dynamic changes of data. How to intelligently judge and remove those noise edges that have a negative impact on specific tasks, and then obtain a base map that is both concise and rich in key information, is one of the challenges facing this field.

[0033] The present specification is described in detail below with reference to specific embodiments.

[0034] See also Figure 1 , is a schematic diagram of a data processing system scenario provided in this specification. Figure 1 As shown, the data processing system may include at least a client cluster and a service platform 100 .

[0035] The client cluster may include at least one client, such as Figure 1 As shown, it specifically includes client 1 corresponding to user 1, client 2 corresponding to user 2, ..., client n corresponding to user n, where n is an integer greater than 0.

[0036] Each client in the client cluster may be an electronic device with communication function, including but not limited to: wearable device, handheld device, personal computer, tablet computer, vehicle-mounted device, smart phone, computing device or other processing device connected to wireless modem, etc. Electronic devices may be called different names in different networks, such as: user equipment, access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, electronic device, wireless communication device, user agent or user device, cellular phone, cordless phone, personal digital assistant (PDA), electronic device in 5G network or future evolution network, etc.

[0037] The service platform 100 can be a separate server device, such as a rack-mounted, blade, tower, or cabinet-mounted server device, or a workstation, mainframe computer, or other hardware device with strong computing power; it can also be a server cluster composed of multiple servers, and the servers in the service cluster can be composed in a symmetrical manner, wherein each server has equivalent functions and status in the transaction link, and each server can provide services to the outside independently, and the independent service can be understood as not requiring the assistance of other servers.

[0038] It should be noted that the service platform 100 and at least one client in the client cluster establish a communication connection through a network for interactive communication, wherein 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 local area network. In one or more embodiments of the specification, technologies and / or formats including Hyper Text Mark-up Language (HTML), Extensible Markup Language (XML), etc. are used to represent data (such as a target compressed package) exchanged through a network. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec), etc. 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 data communication technologies.

[0039] The data processing system embodiments provided in this specification and the data processing methods in one or more embodiments belong to the same concept. The execution subject corresponding to the data processing method involved in one or more embodiments of the specification can be the above-mentioned service platform 100; the execution subject corresponding to the data processing method involved in one or more embodiments of the specification can also be the electronic device corresponding to the client, which is determined based on the actual application environment. The implementation process of the data processing system embodiment can be found in the following method embodiment, which will not be repeated here.

[0040] based on Figure 1 The scene diagram is shown, and the data processing method provided by one or more embodiments of this specification is introduced in detail below.

[0041] See also Figure 2 , is a flow chart of a data processing method for one or more embodiments of this specification, which can be implemented by a computer program and can be run on a data processing device based on the von Neumann system. The computer program can be integrated into an application or run as an independent tool application. The data processing device can be a service platform.

[0042] Specifically, the data processing method includes:

[0043] S102: Acquire first graph structure data corresponding to the target transaction, and determine at least one transaction subgraph structure data corresponding to the first graph structure data;

[0044] It is understandable that in a large number of different types of transaction scenarios, a large amount of structured data and information is collected and accumulated from different data sources, and practical applications involve the need to describe these structured data and information. Based on this, in these actual transaction scenarios, it is often necessary to structure and represent these large amounts of information or even data in the form of graph structure data before storing them.

[0045] It can be understood that in practical applications, graph structure construction technology can be used to record graph structure data in target transactions for corresponding target transactions (such as financial transaction scenarios, shopping transaction scenarios, instant messaging transaction scenarios, content recommendation transaction scenarios, etc.).

[0046] Graph structured data refers to a graph with a large number of nodes and frequent changes, usually containing complex network data of a large number of nodes and edges, such as a graph with more than 5,000 nodes, which changes every minute, and can represent entities and the relationships between them. Graph structured data is widely present in many fields, such as social networks, financial transactions, logistics networks, etc. In some embodiments, graph structured data can also be referred to as large-scale graph structured data, and a node threshold or an edge threshold is usually set. When the nodes in the graph structured data exceed the node threshold and / or the edges exceed the edge threshold, the graph structured data can be regarded as large-scale graph structured data.

[0047] Components of graph structure data:

[0048] Vertices: The basic unit in a graph, representing an entity. For example, in a social network, a node can represent a user; in the financial field, a node can represent a company, an account, or a transaction.

[0049] Edges: Lines connecting nodes, representing relationships or interactions between nodes. For example, in social networks, edges can represent friendships between users; in financial data, edges can represent transaction relationships or capital flows.

[0050] The first graph structure data may be historical graph data or original graph data of the service platform on the target transaction. The historical data or original graph data is used to perform edge structure knowledge distillation with the aid of the first transaction graph processing model to obtain the second transaction graph processing model.

[0051] Considering that the first graph structure data is usually huge, based on this, when constructing the first transaction graph processing model of the target transaction, one or more transaction subgraph structure data are extracted, that is, at least one transaction subgraph structure data corresponding to the first graph structure data is determined;

[0052] Optionally, the first graph structure data may be subgraph sampled by adopting a graph data sampling method, and the original first graph structure data may be sampled into small transaction subgraph structure data;

[0053] S104: determining a first transaction graph processing model based on the transaction subgraph structure data, performing edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model, wherein the second transaction graph processing model is a lightweight model relative to the first transaction graph processing model;

[0054] Schematically, the transaction subgraph structure data is input into the pre-built interpretable first transaction graph processing model. The first transaction graph processing model is usually a graph neural network model. Under different transactions, this specification selects the corresponding first graph structure data under different transactions and inputs them into the first transaction graph processing model for transaction graph processing training. After completing the transaction graph processing training task, the first transaction graph processing model has the edge structure knowledge of the graph structure data under the target transaction. Then, the first transaction graph processing model is subjected to edge structure knowledge distillation processing to obtain a second transaction graph processing model.

[0055] The second transaction graph processing model is a lightweight model relative to the first transaction graph processing model.

[0056] In this specification, the lightweight model can be understood as comparing the second transaction graph processing model with the first transaction graph processing model, and the second transaction graph processing model is a lightweight model relative to the first transaction graph processing model; in terms of processing resource usage, the processing resource usage of the second transaction graph processing model is less than that of the first transaction graph processing model; in terms of computational complexity, the computational complexity of the second transaction graph processing model is less than that of the first transaction graph processing model; in terms of model architecture, the model architecture complexity of the second transaction graph processing model is less than that of the first transaction graph processing model; in terms of storage space, the storage space of the second transaction graph processing model is less than that of the first transaction graph processing model;

[0057] In this specification, (edge ​​structure) knowledge distillation is to transfer the knowledge of a large and complex model (called the first transaction graph processing model) to a smaller and simpler model (the second transaction graph processing model). The purpose of knowledge distillation is usually to obtain a model that is still powerful but more efficient in performance.

[0058] In a feasible implementation, a first transaction graph processing model may be used to parse edge structure information based on the transaction subgraph structure data to obtain edge structure importance information for the target transaction, and the first transaction graph processing model may be subjected to edge structure knowledge distillation processing for the edge structure importance information to obtain a second transaction graph processing model.

[0059] Exemplarily, a first transaction graph processing model is used to perform edge structure information parsing task training based on multiple transaction subgraph structure data. During each round of task training, the edge structure information parsing task is to identify and learn the importance contribution of the edges in each transaction subgraph structure data to the target transaction task (such as the classification task, clustering task, recommendation task, etc. under the target transaction). Each round of task training will output the edge structure importance information for the target transaction, and the edge structure importance information is the importance contribution of the edges in the transaction subgraph structure data to the target transaction task. After completing the edge structure information parsing task training, the first transaction graph processing model is subjected to edge structure knowledge distillation processing on the edge structure importance information to obtain a second transaction graph processing model.

[0060] Optionally, the first transaction graph processing model may be a model such as a graph neural network (GNN), a graph convolutional network (GCN), a graph attention network (GAT), etc.;

[0061] Optionally, the second transaction graph processing model is a lightweight model relative to the first transaction graph processing model, which means that after knowledge distillation or other optimization methods, it has smaller computational complexity and storage requirements, but can still effectively process graph structure data. It has one or more characteristics such as fewer parameters, efficient reasoning, scalability, and easy maintenance. The second transaction graph processing model can select a lightweight processing model architecture, such as a lightweight graph neural network, a multi-layer perceptron (MLP) model, and other models;

[0062] In a feasible implementation, the first transaction graph processing model includes a transaction graph processing model created based on a graph neural network, and the second transaction graph processing model includes a transaction graph processing model created based on a plurality of perceptron networks. By using an interpretable graph neural network, the importance of each subgraph structure data or edge in the graph structure data is learned through training, and the knowledge distillation technology is adopted to effectively transfer the knowledge containing the edge importance probability information to the second transaction graph processing model with lower computational cost. Filtering is performed through the second transaction graph processing model.

[0063] S106: Acquire the second graph structure data of the target transaction, and perform graph sparse processing on the second graph structure data based on the second transaction graph processing model to obtain the target transaction graph structure data;

[0064] Exemplarily, the second graph structure data may be online graph structure data updated in real time by the target transaction. The second transaction graph processing obtained above may then be used to perform graph sparsification processing on the second graph structure data that is quickly updated by the target transaction to generate a streamlined base map reflecting its core structure, i.e., the target transaction graph structure data. Based on the previously trained second transaction graph processing model, this online graph structure data is subjected to graph sparsification processing. Graph sparsification processing aims to reduce non-core edges in the graph structure data while retaining key information, thereby generating a more streamlined target transaction graph structure data. This step significantly improves the efficiency of subsequent analysis and processing, thereby achieving efficient and robust graph processing and reasoning performance;

[0065] like Figure 3 As shown, Figure 3 It is a scene diagram of graph structure data processing. Figure 3 In the example, the first graph structure data under a target transaction is sampled to obtain a number of transaction subgraph structure data. The first graph structure data may be the historical graph data or original graph data of the service platform on the target transaction. The transaction subgraph structure data is input into the pre-constructed interpretable first transaction graph processing model (such as Figure 3 The first transaction graph processing model is usually a graph neural network model. In different transactions, this specification selects the corresponding first graph structure data under different transactions and inputs them into the first transaction graph processing model for transaction graph processing training. After completing the transaction graph processing training task, the first transaction graph processing model has the edge structure knowledge of the graph structure data under the target transaction. Then, based on the selected second transaction graph processing model (such as Figure 3 As shown in the figure, a multi-layer perceptron (MLP) model is used to perform edge structure knowledge distillation processing using the first transaction graph processing model to obtain a trained second transaction graph processing model. Based on the previously trained second transaction graph processing model, the online graph structure data is subjected to graph sparse processing. Graph sparse processing aims to reduce non-core edges in the graph structure data while retaining key information, thereby generating a more streamlined target transaction graph structure data.

[0066] In this specification, by introducing edge structure knowledge distillation in the graph structure data processing scenario, the obtained small model (second transaction graph processing model) not only carries the core recognition capability of the interpretable first transaction graph processing model, but also greatly speeds up the processing speed of the second graph structure data that is quickly updated for the target transaction due to its smaller scale. This enables the service platform to respond quickly to data changes even in a resource-constrained environment, ensuring the real-time update and deployment of the target transaction model (content recommendation, credit scoring, financial risk control and other models). This graph sparsification method reduces the computational complexity while maintaining the capture of key information of the original base map, which has important application value in scientific research scenarios such as financial risk control, social network analysis, and even bioinformatics.

[0067] S108: Perform target transaction processing based on the target transaction graph structure data.

[0068] Exemplarily, the target transaction graph structure data that has been subjected to sparse processing is used. The target transaction graph structure data has removed unnecessary edges, retains core relationships, and is suitable for subsequent analysis and processing of target transactions.

[0069] Optionally, a target transaction knowledge graph for the target transaction may be constructed using knowledge graph technology based on the target transaction graph structure data.

[0070] Optionally, target transaction processing may be performed on the target transaction graph structure data based on specific target transaction tasks. For example, target transaction tasks include but are not limited to risk assessment tasks, user behavior analysis tasks, recommendation generation tasks, network communication analysis tasks, and the like.

[0071] For example, in a financial scenario: when processing financial transaction data, potential fraudulent transactions are identified through the target transaction graph structure data, and risk control measures are triggered based on the analysis results.

[0072] For example, in a social media scenario: using the interaction graph between users in the target transaction graph structure data, analyze which users have a strong influence, so as to formulate corresponding promotion strategies.

[0073] In one or more embodiments of the present specification, first graph structure data corresponding to the target transaction is obtained, transaction subgraph structure data corresponding to the first graph structure data is determined, a first transaction graph processing model is determined based on the transaction subgraph structure data, edge structure knowledge distillation is performed on the first transaction graph processing model to obtain a second transaction graph processing model, and graph sparsification is performed on the second graph structure data of the target transaction based on the second transaction graph processing model to obtain target transaction graph structure data. By combining the expressiveness of the first graph structure data of the target transaction through deep learning and the computational efficiency of the lightweight second transaction graph processing model, the contradiction between computing performance and model accuracy of large-scale graph sparsification is resolved. Subsequently, target transaction processing can be performed based on the target transaction graph structure data, showing significant effects in improving model performance, increasing reasoning speed, and reducing computing resource consumption, which helps to promote technological innovation and commercial applications in related transaction scenarios.

[0074] See also Figure 4 , Figure 4 This is a flow chart of an edge structure knowledge distillation method proposed in one or more embodiments of this specification. The following method can be used to perform the edge structure information parsing based on the transaction subgraph structure data using the first transaction graph processing model to obtain the edge structure importance information for the target transaction, and perform the edge structure knowledge distillation processing on the edge structure importance information on the first transaction graph processing model to obtain the second transaction graph processing model:

[0075] S202: using all the transaction subgraph structure data to perform at least one round of model training on the first transaction graph processing model, and performing graph structure analysis processing on the transaction subgraph structure data by the first transaction graph processing model to obtain edge structure importance information;

[0076] Exemplarily, a first transaction graph processing model for a target transaction is created. Multiple transaction subgraph structure data can be used to perform edge structure information parsing task training on the first transaction graph processing model. During each round of task training, the edge structure information parsing task is to identify and learn the importance contribution of the edges in each transaction subgraph structure data to the target transaction task (such as the classification task, clustering task, recommendation task, etc. under the target transaction). Each round of task training will output the edge structure importance information for the target transaction. The edge structure importance information is also the importance contribution of the edges in the transaction subgraph structure data to the target transaction task. After completing the edge structure information parsing task training, the first transaction graph processing model adapted to the target transaction is obtained.

[0077] For example, please refer to Figure 5 , Figure 5 It is a flowchart of the first transaction graph processing model, specifically:

[0078] S3002: Create an initial first transaction graph processing model;

[0079] Optionally, one or more graph processing models such as a graph neural network (GNN), a graph convolutional network (GCN), a graph attention network (GAT), etc. may be selected to create an initial first transaction graph processing model;

[0080] S3004: inputting all the transaction subgraph structure data into the initial first transaction graph processing model to perform at least one round of model training until the model training is completed to obtain the first transaction graph processing model;

[0081] Exemplarily, all transaction subgraph structure data are used as sample data, and expert-side services can be used to annotate edge structure importance information labels corresponding to the transaction subgraph structure data in advance;

[0082] The model input of the (initial) first transaction graph processing model is the transaction subgraph structure data; the (initial) first transaction graph processing model is used to identify the importance contribution of the edge in the transaction subgraph structure data to the target transaction task (such as the classification task, clustering task, recommendation task, etc. under the target transaction), and output the predicted edge structure importance information;

[0083] In each round of model training, the predicted edge structure importance information output by the (initial) first transaction graph processing model is determined, the first model loss is calculated using a first loss function based on the predicted edge structure importance information and the edge structure importance information label, and the model parameters of the (initial) first transaction graph processing model are adjusted using the first model loss until the (initial) first transaction graph processing model meets the model training end condition, thereby obtaining the first transaction graph processing model;

[0084] Optionally, the model training end condition may include, for example, the first model loss 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 end condition can be determined based on actual conditions and is not specifically limited here.

[0085] Optionally, the first loss function may adopt one or more loss functions such as cross entropy loss, hinge loss, Euclidean distance loss, etc. in related technologies.

[0086] S3006: Determine edge structure importance information for the transaction subgraph structure data based on the first transaction graph processing model.

[0087] In an illustrative manner, after the model is extracted from the trained first transaction graph processing model, for the edge structure importance information of the transaction subgraph structure data, all the transaction subgraph structure data can be re-input into the first transaction graph processing model to output the edge structure importance information of each transaction subgraph structure data;

[0088] S204: extracting at least one edge structure data corresponding to the edge structure importance information from the transaction subgraph structure data;

[0089] The edge structure transaction information is data describing the importance and contribution of a "certain edge" in the transaction subgraph structure data to the target transaction task (such as the classification task, clustering task, recommendation task, etc. under the target transaction), while the edge structure data is the edge structure description information describing the "certain edge" corresponding to the edge structure transaction information. The edge structure description information can be understood as the meta-information of the edge.

[0090] In this specification, at least one edge structure data corresponding to the edge structure importance information is extracted from the transaction subgraph structure data to construct the model input data of the second transaction graph processing model in the subsequent knowledge migration stage, and the edge structure importance information corresponding to the edge architecture data is used as the output label data.

[0091] S206: Using the edge structure data and the edge structure importance information, perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model.

[0092] The edge structure data is used as the model input data for the (initial) second transaction graph processing model in the knowledge distillation processing stage, and the edge structure importance information is used as the output label data;

[0093] By using the edge structure data and the edge structure importance information to perform edge structure knowledge distillation processing on the first transaction graph processing model, a second transaction graph processing model can be obtained.

[0094] In one or more embodiments of the present specification, by introducing edge structure knowledge distillation in the graph structure data processing scenario, the obtained small model (second transaction graph processing model) not only carries the core recognition capability of the interpretable first transaction graph processing model, but also greatly speeds up the processing speed of the second graph structure data that is quickly updated for the target transaction due to its smaller scale. This enables the service platform to respond quickly to data changes even in a resource-constrained environment, ensuring the real-time update and deployment of the target transaction model (content recommendation, credit scoring, financial risk control and other models). This graph sparsification method reduces the computational complexity while maintaining the capture of key information of the original base map, which has important application value in scientific research scenarios such as financial risk control, social network analysis, and even bioinformatics.

[0095] For example, please refer to Figure 6 , Figure 6 It is a flowchart of the second transaction graph processing model, specifically:

[0096] S4002: creating an initial second transaction graph processing model, acquiring model edge structure knowledge from the first transaction graph processing model, and migrating the model edge structure knowledge to the initial second transaction graph processing model;

[0097] Optionally, the (initial) second transaction graph processing model is a lightweight model relative to the first transaction graph processing model, which means that after knowledge distillation or other optimization methods, it has smaller computational complexity and storage requirements, but can still effectively process graph structure data. It has one or more characteristics such as fewer parameters, efficient reasoning, scalability, and easy maintenance.

[0098] Optionally, the (initial) second transaction graph processing model may select a lightweight processing model architecture, such as a lightweight graph neural network, a multi-layer perceptron (MLP) model, and the like;

[0099] Exemplarily, the model edge structure knowledge can be obtained from the first transaction graph processing model, and the model edge structure knowledge is stored in the model network layer parameters of the first transaction graph processing model, that is, the model network layer parameters can be extracted and applied to the initial second transaction graph processing model to initialize the model parameters, thereby completing the migration of the model edge structure knowledge. The weights or parameters of the (initial) second transaction graph processing model are initialized with edge structure importance information, so that it can quickly learn the characteristics of important edges.

[0100] S4004: labeling the edge structure data with an edge structure importance label based on the edge structure importance information;

[0101] Exemplarily, the edge structure importance information determined by the first transaction graph processing model will be used as a label of the edge structure data, and the label is also the edge structure importance label;

[0102] S4006: Inputting the edge structure data into an initial second transaction graph processing model for at least one round of model training, and identifying predicted edge structure importance information of the edge structure data for the target transaction through the initial second transaction graph processing model during the model training;

[0103] The edge structure data is trained through the initial second transaction graph processing model. During the model forward propagation training process, the initial second transaction graph processing model learns to predict the importance related to the target transaction based on the input edge structure data, and outputs the predicted edge structure importance information.

[0104] S4008: Determine a target model loss based on the predicted edge structure importance information and the edge structure importance label, and use the target model loss to adjust model parameters of the initial second transaction graph processing model until model training is completed to obtain a second transaction graph processing model.

[0105] The target model loss is calculated using the corresponding loss function, which uses the predicted edge structure importance information and the actual edge structure importance label to calculate the target model loss. For example, a suitable loss function, such as mean square error or cross entropy loss, can be selected in advance to reflect the gap between the prediction and the actual.

[0106] The target model loss is used to adjust the model parameters of the initial second transaction graph processing model. The target model loss is used to adjust the parameters of the initial second transaction graph processing model through a back propagation optimization algorithm (such as Adam or SGD) to reduce the loss and optimize the model performance. The training is iterative and the process is repeated until the loss function converges and the model reaches the expected performance level, thereby obtaining the final second transaction graph processing model.

[0107] In one or more embodiments of this specification, the training process of the initial second transaction graph processing model is optimized by transferring edge structure knowledge and annotating importance. By continuously adjusting model parameters, a lightweight model with better performance on target transactions is eventually obtained. This process not only improves the accuracy of the model, but also increases its sensitivity to edge structure information, thereby better coping with the complexity in practical applications.

[0108] Optional, for reference Figure 7 , Figure 7 The invention is a flowchart of graph sparse processing. Specifically, the graph sparse processing of the second graph structure data based on the second transaction graph processing model to obtain the target transaction graph structure data can be performed by referring to the following method:

[0109] S5002: Based on the second graph structure data, determine target edge structure importance information corresponding to at least one group of edge structures through the second transaction graph processing model;

[0110] After obtaining the second graph structure data of the target transaction, the second graph structure data is input into the second transaction graph processing model. For example, the daily second graph structure data of the target transaction can be input into the second transaction graph processing model. The second graph structure data includes a large number of edge structures. The edge structure information parsing task is processed by the second transaction graph processing model. The structure information parsing task is to identify the importance contribution of the edge in the second graph structure data to the target transaction task (such as the classification task, clustering task, recommendation task, etc. under the target transaction), thereby generating the target edge structure importance information;

[0111] S5004: Performing graph structure filtering processing in the second graph structure data based on the target edge structure importance information to obtain target transaction graph structure data:

[0112] For example, the following method may be referred to:

[0113] B2: inputting the second graph structure data into the second transaction graph processing model, and determining a target edge structure that does not match the edge structure importance threshold information in the second graph structure data through the second transaction graph processing model;

[0114] The edge structure importance threshold information includes a threshold set for the edge structure importance degree. If the value of the edge structure importance information is greater than the threshold in the edge structure importance threshold information, the edge structure is considered important for the target transaction task. If the value of the edge structure importance information is less than or equal to the threshold in the edge structure importance threshold information, the edge structure is considered unimportant for the target transaction task.

[0115] The second graph structure data is input into a second transaction graph processing model, and the second transaction graph processing model can determine target edge structure importance information corresponding to each edge structure, and then compare the target edge structure importance information with the edge structure importance threshold information, and determine that the target edge structure corresponding to the threshold value in the edge structure importance threshold information is less than or equal to the value of the edge structure importance information, and the target edge structure belongs to the edge structure importance threshold information mismatch;

[0116] B4: Determine a target graph structure corresponding to the target edge structure in the second graph structure data, where the target graph structure includes a target edge and / or a target node;

[0117] Determine a target graph structure corresponding to the target edge structure in the second graph structure data, where the target graph result includes at least one of a target edge and a target node;

[0118] B6: Filter out the target graph structure from the second graph structure data to obtain target transaction graph structure data.

[0119] All target graph structures are filtered out in the second graph structure data to retain core relationships, and graph sparsification processing is completed to obtain target transaction graph structure data.

[0120] In this specification, the second graph structure data is used to perform graph sparsification, which has better effects than sparsification techniques such as edge sampling, node sampling, and community detection in related technologies. The complexity of the graph is reduced by extracting representative subgraphs, reducing the number of nodes in the graph, or clustering the graph into closely connected communities. In related technologies, graph compression may oversimplify the structure of the graph, lose important information, and often requires manual setting of indicators or reliance on specific heuristic rules, and lacks task relevance, which means that the most valuable information may not be retained in specific application tasks. The method of one or more embodiments of this specification relies on simple heuristic rules, but instead uses interpretable first graph structure data to learn the importance of each edge in the graph, ensuring that the selection of edges is closely related to specific tasks. Through knowledge distillation, edge importance decisions are passed to the second graph structure data with lower computational cost, which makes the processing fast and can retain edges that are valuable to specific tasks, efficient and accurate;

[0121] B8: monitoring graph structure update data for the second graph structure data, where the graph structure update data includes graph structure update edge information and / or graph structure update node information;

[0122] B10: Input the graph structure update data into the second transaction graph processing model, and execute the step of determining the target edge structure that does not match the edge structure importance threshold information in the second graph structure data through the second transaction graph processing model.

[0123] In this specification, the second transaction graph processing model supports dynamic graph data updates relative to the second graph structure data. For a dynamically changing graph, the graph structure update data of the target transaction for the second graph structure data can be dynamically monitored, and then the graph structure update data can be dynamically input into the second transaction graph processing model to execute the target edge structure that does not match the edge structure importance threshold information in the second graph structure data determined by the second transaction graph processing model, and the target graph structure is filtered out in the second graph structure data to obtain the target transaction graph structure data.

[0124] The second transaction graph processing model can continuously adapt to the newly generated graph structure update data, where the graph structure update data includes edge change update data and node change update data, to ensure that the sparse graph (target transaction graph structure data) always reflects the latest data structure.

[0125] In one or more embodiments of this specification, the following technical effects can be achieved:

[0126] 1. Edge importance learning based on the first transaction graph processing model (such as graph neural network GNN): This solution learns and understands the importance of each subgraph or edge in a large-scale graph through an interpretable model. This method can intelligently identify and maintain relationships that are beneficial to the final task, rather than relying solely on predetermined rules or thresholds.

[0127] 2. Application of knowledge distillation technology in graph sparsification: Innovatively use the knowledge distillation process to transfer the edge relationship importance judgment probability information in the complex first transaction graph processing model to the second transaction graph processing model with lower computational cost (such as the multi-layer perceptron (MLP) model), which reduces the resource consumption and processing time of graph sparsification tasks on large-scale graphs.

[0128] 3. Balance between efficiency and accuracy: Through knowledge distillation, this solution not only maintains the fine-grained graph structure information learned by the complex first transaction graph processing model, but also enables small models (such as the second transaction graph processing model) to quickly and accurately perform graph sparsification while greatly reducing the computational burden.

[0129] 4. Adapt to changes in dynamic graph data: This solution is particularly suitable for dynamic graph data and can effectively adapt to graph structures that change over time, such as the addition or deletion of nodes and edges. It supports real-time updates and adaptation of the system without frequent retraining.

[0130] 5. Improve reasoning speed and reduce computing resource consumption: By reducing the size and complexity of the graph, this solution significantly improves the reasoning speed of the model while reducing the demand for storage and computing resources.

[0131] 6. Versatility and flexibility: This sparsification method is not only practical in the fields of finance and social networks, but also shows its wide application potential in other business scenarios such as recommendation systems and bioinformatics, enhancing the practicality and universality of the solution.

[0132] The following will be combined Figure 8 , the data processing device provided in this specification is introduced in detail. It should be noted that, Figure 8 The data processing device shown is used to execute the Figure 1 to Figure 7 For the convenience of explanation, only the parts related to this specification are shown. For specific technical details not disclosed, please refer to this specification. Figure 1 to Figure 7 The embodiment shown.

[0133] See also Figure 8, which shows a schematic diagram of the structure of the data processing device of this specification. The data processing device 1 can be implemented as all or part of the device through software, hardware or a combination of both. According to some embodiments, the data processing device 1 includes a data acquisition module 11, a knowledge distillation module 12, a graph sparsification module 13 and a transaction processing module 14, which are specifically used to:

[0134] A data acquisition module 11 is used to acquire first graph structure data corresponding to a target transaction, and determine at least one transaction subgraph structure data corresponding to the first graph structure data;

[0135] A knowledge distillation module 12 is used to determine a first transaction graph processing model based on the transaction subgraph structure data, and to perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model, wherein the second transaction graph processing model is a lightweight model relative to the first transaction graph processing model;

[0136] A graph sparsification module 13, configured to obtain the second graph structure data of the target transaction, and perform graph sparsification processing on the second graph structure data based on the second transaction graph processing model to obtain the target transaction graph structure data;

[0137] The transaction processing module 14 is used to perform target transaction processing based on the target transaction graph structure data.

[0138] Optionally, the knowledge distillation module 12 is used to:

[0139] Based on the transaction subgraph structure data, the first transaction graph processing model is used to parse the edge structure information to obtain edge structure importance information for the target transaction;

[0140] The first transaction graph processing model is subjected to edge structure knowledge distillation processing for the edge structure importance information to obtain a second transaction graph processing model.

[0141] Optionally, the knowledge distillation module 12 is used to:

[0142] Using all of the transaction subgraph structure data to perform at least one round of model training on the first transaction graph processing model, and performing graph structure analysis processing on the transaction subgraph structure data through the first transaction graph processing model to obtain edge structure importance information;

[0143] Extracting at least one edge structure data corresponding to the edge structure importance information from the transaction subgraph structure data;

[0144] The edge structure data and the edge structure importance information are used to perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model.

[0145] Optionally, the knowledge distillation module 12 is used to:

[0146] Creating an initial first transaction graph processing model;

[0147] Inputting all the transaction subgraph structure data into the initial first transaction graph processing model to perform at least one round of model training until the model training is completed to obtain the first transaction graph processing model;

[0148] Based on the first transaction graph processing model, edge structure importance information for the transaction subgraph structure data is determined.

[0149] Optionally, the knowledge distillation module 12 is used to:

[0150] Creating an initial second transaction graph processing model, acquiring model edge structure knowledge from the first transaction graph processing model, and migrating the model edge structure knowledge to the initial second transaction graph processing model;

[0151] labeling the edge structure data with an edge structure importance label based on the edge structure importance information;

[0152] Inputting the edge structure data into an initial second transaction graph processing model for at least one round of model training, and identifying predicted edge structure importance information of the edge structure data for the target transaction through the initial second transaction graph processing model during the model training;

[0153] A target model loss is determined based on the predicted edge structure importance information and the edge structure importance label, and the target model loss is used to adjust model parameters of the initial second transaction graph processing model until model training is completed to obtain a second transaction graph processing model.

[0154] Optionally, the first transaction graph processing model includes a transaction graph processing model created based on a graph neural network, and the second transaction graph processing model includes a transaction graph processing model created based on a multi-site perceptron network.

[0155] Optionally, the graph sparsification module 13 is used to:

[0156] Based on the second graph structure data, target edge structure importance information corresponding to at least one group of edge structures is determined through the second transaction graph processing model, and graph structure filtering processing is performed in the second graph structure data based on the target edge structure importance information to obtain target transaction graph structure data.

[0157] Optionally, the graph sparsification module 13 is used to: input the second graph structure data into the second transaction graph processing model, and determine the target edge structure that does not match the edge structure importance threshold information in the second graph structure data through the second transaction graph processing model;

[0158] Determine a target graph structure corresponding to the target edge structure in the second graph structure data, wherein the target graph structure includes a target edge and / or a target node;

[0159] The target graph structure is filtered out from the second graph structure data to obtain target transaction graph structure data.

[0160] Optionally, the graph sparsification module 13 is used to: monitor graph structure update data for the second graph structure data, wherein the graph structure update data includes graph structure update edge information and / or graph structure update node information;

[0161] The graph structure update data is input into the second transaction graph processing model, and the step of determining the target edge structure that does not match the edge structure importance threshold information in the second graph structure data through the second transaction graph processing model is performed.

[0162] It should be noted that, when the data processing device provided in the above embodiment executes the data processing method, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the data processing device provided in the above embodiment and the data processing method embodiment belong to the same concept, and the implementation process thereof is detailed in the method embodiment, which will not be repeated here.

[0163] The above serial numbers in this specification are for description only and do not represent the advantages or disadvantages of the embodiments.

[0164] In one or more embodiments of the present specification, first graph structure data corresponding to the target transaction is obtained, transaction subgraph structure data corresponding to the first graph structure data is determined, a first transaction graph processing model is determined based on the transaction subgraph structure data, edge structure knowledge distillation is performed on the first transaction graph processing model to obtain a second transaction graph processing model, and graph sparsification is performed on the second graph structure data of the target transaction based on the second transaction graph processing model to obtain target transaction graph structure data. By combining the expressiveness of the first graph structure data of the target transaction through deep learning and the computational efficiency of the lightweight second transaction graph processing model, the contradiction between computing performance and model accuracy of large-scale graph sparsification is resolved. Subsequently, target transaction processing can be performed based on the target transaction graph structure data, showing significant effects in improving model performance, increasing reasoning speed, and reducing computing resource consumption, which helps to promote technological innovation and commercial applications in related transaction scenarios.

[0165] The present specification also provides a computer storage medium, which can store multiple instructions, which are suitable for being loaded and executed by a processor as described above. Figure 1 to Figure 7The data processing method of the embodiment shown in the figure can be specifically executed by referring to Figure 1 to Figure 7 The specific description of the illustrated embodiment will not be repeated here.

[0166] The present 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 as described above. Figure 1 to Figure 7 The data processing method of the embodiment shown in the figure can be specifically executed by referring to Figure 1 to Figure 7 The specific description of the illustrated embodiment will not be repeated here.

[0167] Please refer to Fig. 9 , is a block diagram of a structure of an electronic device provided in an embodiment of this specification. The electronic device in this specification may include one or more of the following components: a processor 1010, a memory 1020, an input device 1030, an output device 1040, and a bus 1050. The processor 1010, the memory 1020, the input device 1030, and the output device 1040 may be connected via a bus 1050.

[0168] The processor 1010 may include one or more processing cores. The processor 1010 uses various interfaces and lines to connect various parts of the entire electronic device, and executes various functions and processes data of the electronic device 100 by running or executing instructions, programs, code sets or instruction sets stored in the memory 1020, and calling data stored in the memory 1020. Optionally, the processor 1010 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 1010 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 1010, but may be implemented separately through a communication chip.

[0169] The memory 1020 may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory 1020 includes a non-transitory computer-readable storage medium. The memory 1020 may be used to store instructions, programs, codes, code sets, or instruction sets.

[0170] The input device 1030 is used to receive input instructions or data, and the input device 1030 includes but is not limited to a keyboard, a mouse, a camera, a microphone or a touch device. The output device 1040 is used to output instructions or data, and the output device 1040 includes but is not limited to a display device and a speaker. In the embodiment of this specification, the input device 1030 can be a temperature sensor for obtaining the operating temperature of the electronic device. The output device 1040 can be a speaker for outputting audio signals.

[0171] In addition, those skilled in the art will appreciate that the structure of the electronic device shown in the above drawings does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown, or combine certain components, or arrange the components differently. 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 described in detail here.

[0172] In the embodiments of this specification, the execution subject of each step may be the electronic device described above. Optionally, the execution subject of each step is the operating system of the electronic device. The operating system may be an Android system, an IOS system, or other operating systems, which is not limited in the embodiments of this specification.

[0173] exist Fig. 9 In the electronic device, the processor 1010 can be used to call the program stored in the memory 1020 and execute it to implement the data processing method as described in the various method embodiments of this specification.

[0174] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0175] 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 relevant laws, regulations and standards of relevant countries and regions. For example, the graph structure data involved in this specification are all obtained with full authorization.

[0176] The above disclosure is only the preferred embodiment of this specification, which certainly cannot be used to limit the scope of rights of this specification. Therefore, equivalent changes made according to the claims of this specification are still within the scope covered by this specification.

Claims

1. A data processing method, the method comprising: Acquire first graph structure data corresponding to the target transaction, and determine at least one transaction subgraph structure data corresponding to the first graph structure data; Determine a first transaction graph processing model based on the transaction subgraph structure data, perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model, wherein the second transaction graph processing model is a lightweight model relative to the first transaction graph processing model; Acquire the second graph structure data of the target transaction, and perform graph sparse processing on the second graph structure data based on the second transaction graph processing model to obtain the target transaction graph structure data; Target transaction processing is performed based on the target transaction graph structure data.

2. The method according to claim 1, wherein determining a first transaction graph processing model based on the transaction subgraph structure data, and performing edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model comprises: Based on the transaction subgraph structure data, the first transaction graph processing model is used to parse the edge structure information to obtain edge structure importance information for the target transaction; The first transaction graph processing model is subjected to edge structure knowledge distillation processing for the edge structure importance information to obtain a second transaction graph processing model.

3. The method according to claim 2, wherein the first transaction graph processing model is used to parse the edge structure information based on the transaction subgraph structure data to obtain the edge structure importance information for the target transaction, and the first transaction graph processing model is subjected to edge structure knowledge distillation processing for the edge structure importance information to obtain the second transaction graph processing model, comprising: Using all of the transaction subgraph structure data to perform at least one round of model training on the first transaction graph processing model, and performing graph structure analysis processing on the transaction subgraph structure data through the first transaction graph processing model to obtain edge structure importance information; Extracting at least one edge structure data corresponding to the edge structure importance information from the transaction subgraph structure data; The edge structure data and the edge structure importance information are used to perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model.

4. The method according to claim 3, wherein the first transaction graph processing model is trained for at least one round using all the transaction subgraph structure data, and edge structure importance information is obtained by performing graph structure analysis processing on the transaction subgraph structure data using the first transaction graph processing model, including: Creating an initial first transaction graph processing model; Inputting all the transaction subgraph structure data into the initial first transaction graph processing model to perform at least one round of model training until the model training is completed to obtain the first transaction graph processing model; Based on the first transaction graph processing model, edge structure importance information for the transaction subgraph structure data is determined.

5. The method according to claim 3, wherein the step of using the edge structure data and the edge structure importance information to perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model comprises: Creating an initial second transaction graph processing model, acquiring model edge structure knowledge from the first transaction graph processing model, and migrating the model edge structure knowledge to the initial second transaction graph processing model; labeling the edge structure data with an edge structure importance label based on the edge structure importance information; Inputting the edge structure data into an initial second transaction graph processing model for at least one round of model training, and identifying predicted edge structure importance information of the edge structure data for the target transaction through the initial second transaction graph processing model during the model training; A target model loss is determined based on the predicted edge structure importance information and the edge structure importance label, and the target model loss is used to adjust model parameters of the initial second transaction graph processing model until model training is completed to obtain a second transaction graph processing model.

6. According to the method of claim 1, the first transaction graph processing model includes a transaction graph processing model created based on a graph neural network, and the second transaction graph processing model includes a transaction graph processing model created based on a multi-site perceptron network.

7. According to the method of claim 1, the step of performing graph sparse processing on the second graph structure data based on the second transaction graph processing model to obtain target transaction graph structure data comprises: Based on the second graph structure data, target edge structure importance information corresponding to at least one group of edge structures is determined through the second transaction graph processing model, and graph structure filtering processing is performed in the second graph structure data based on the target edge structure importance information to obtain target transaction graph structure data.

8. The method according to claim 7, wherein the determining, based on the second graph structure data, target edge structure importance information corresponding to at least one group of edge structures through the second transaction graph processing model comprises: Inputting the second graph structure data into the second transaction graph processing model, and determining, through the second transaction graph processing model, a target edge structure that does not match the edge structure importance threshold information in the second graph structure data; Determine a target graph structure corresponding to the target edge structure in the second graph structure data, wherein the target graph structure includes a target edge and / or a target node; The target graph structure is filtered out from the second graph structure data to obtain target transaction graph structure data.

9. The method according to claim 8, further comprising: Monitoring graph structure update data for the second graph structure data, wherein the graph structure update data includes graph structure update edge information and / or graph structure update node information; The graph structure update data is input into the second transaction graph processing model, and the step of determining the target edge structure that does not match the edge structure importance threshold information in the second graph structure data through the second transaction graph processing model is performed.

10. A data processing device, comprising: A data acquisition module, used to acquire first graph structure data corresponding to a target transaction, and determine at least one transaction subgraph structure data corresponding to the first graph structure data; a knowledge distillation module, configured to determine a first transaction graph processing model based on the transaction subgraph structure data, and perform edge structure knowledge distillation processing on the first transaction graph processing model to obtain a second transaction graph processing model, wherein the second transaction graph processing model is a lightweight model relative to the first transaction graph processing model; A graph sparsification module, used for acquiring the second graph structure data of the target transaction, and performing graph sparsification processing on the second graph structure data based on the second transaction graph processing model to obtain the target transaction graph structure data; A transaction processing module is used to perform target transaction processing based on the target transaction graph structure data.

11. 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 according to any one of claims 1 to 9.

12. A computer program product, the computer program product storing 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.

13. 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.

Images