CN-121980592-A - Electronic commerce data security management method and system based on big data

CN121980592ACN 121980592 ACN121980592 ACN 121980592ACN-121980592-A

Abstract

The invention relates to the technical field of data security management, in particular to a method and a system for managing electronic commerce data security based on big data, wherein the method comprises the steps of responding to a preset response mechanism to acquire the association relationship between node security state data of each level node of an electronic commerce supply link and the nodes, and preprocessing to obtain node risk characteristics, association relationship characteristics and data sensitivity characteristics; the method comprises the steps of constructing a chain type association graph according to association relation characteristics of nodes and the nodes, quantitatively calculating to obtain risk conduction parameters, inputting the node risk characteristics into a node risk prediction model to obtain node risk prediction results, inputting the chain type association graph, the risk conduction parameters and the node risk prediction results into the chain type risk conduction prediction model to obtain link conduction risk prediction results, formulating a grading blocking strategy according to the comprehensive node risks, the link conduction risks and data sensitivity characteristics, and performing differentiated management and control on a supply link according to the grading blocking strategy to realize electronic commerce supply chain risk identification and grading safety management and control.

Inventors

WANG GUANHONG
YIN PENG
LI JIANGHONG

Assignees

深圳迅销科技股份有限公司

Dates

Publication Date: 20260505
Application Date: 20260403

Claims (10)

1. The electronic commerce data safety management method based on big data is characterized by comprising the following steps: responding to a preset response mechanism, collecting node security state data and association relation among nodes of each level of the e-commerce supply link, and preprocessing to obtain node risk characteristics, association relation characteristics and data sensitivity characteristics of the nodes of each level; calculating the chain type association map based on a preset quantitative calculation rule to obtain risk conduction parameters, wherein the risk conduction parameters comprise node risk conduction coefficients, side risk transfer probability and side association strength; Inputting node risk characteristics into a pre-trained node risk prediction model, and outputting a node risk prediction result of a single node, wherein the node risk prediction result comprises a node risk level, a risk occurrence probability and a risk type; Inputting a chain type association map, risk conduction parameters and node risk prediction results into a pre-trained chain type risk conduction prediction model, and outputting a link conduction risk prediction result, wherein the link conduction risk prediction result comprises risk conduction probability, a risk conduction path, a risk influence range and a risk burst time window; and according to the node risk prediction result, the link conduction risk prediction result and the data sensitivity characteristic, a grading blocking strategy is formulated, and corresponding differentiated management and control are executed on the supply link according to the grading blocking strategy.
2. The electronic commerce data security management method based on big data according to claim 1, wherein the triggering of the preset response mechanism specifically comprises the following steps: dividing the e-commerce supply link nodes into core hub level nodes, primary conduction level nodes and tail end association level nodes with successively decreasing triggering priorities based on the initially constructed chain association map; according to node risk conduction coefficients and side association intensities obtained by initial calculation of nodes of each level, different trigger thresholds are configured for the nodes of each level, wherein the higher the node risk conduction coefficients and the side association intensities are, the lower the corresponding trigger thresholds are; When the node security state data corresponding to any level node reaches a trigger threshold or the edge risk transfer probability of the level node associated edge reaches a preset conduction threshold, a response mechanism is triggered.
3. The method for managing electronic commerce data security based on big data as claimed in claim 1, wherein the step of constructing a chain type association graph by using node risk characteristics as graph node attributes and association relationship characteristics as association edge attributes specifically comprises the following steps: Carrying out structured embedding processing on node risk characteristics of nodes of each level to form node attribute vectors, wherein the node attribute vectors comprise node risk level weights, risk type identifiers and level labels; Classifying and quantifying the association relation features to form an association side attribute vector, wherein the association side attribute vector comprises association relation types, interaction frequencies, data flow orders and duration; And taking the core hub level node as a topology center, performing topology mapping on the node attribute vector and the associated edge attribute vector according to the level chain type linkage relation among the core hub level node, the primary conduction level node and the terminal associated level node, and constructing a chain type association map comprising the node topology relation.
4. The method for managing electronic commerce data security based on big data as described in claim 3, wherein the step of calculating the chain association map based on a preset quantitative calculation rule to obtain risk conduction parameters comprises the steps of: Determining basic conduction coefficients of nodes of each level according to the belonging level labels in the node attribute vectors, and carrying out weighted correction on the basic conduction coefficients of the nodes of each level through the node risk level weights to obtain node risk conduction coefficients of the nodes of each level; calculating the initial association strength of the association edge based on the association relation type, the interaction frequency, the data flow magnitude and the duration in the association edge attribute vector, and correcting the initial association strength by combining the level matching degree of the nodes at the two ends of the association edge to obtain the edge association strength; And carrying out weighted operation on the node risk conduction coefficient and the side association strength to obtain the side risk transfer probability of the corresponding association side.
5. The method for managing electronic commerce data security based on big data according to claim 3, wherein the step of inputting node risk characteristics into a pre-trained node risk prediction model and outputting a node risk prediction result of a single node specifically comprises the steps of: Carrying out feature fusion on node risk features of nodes of each level and corresponding node attribute vectors to obtain fused node risk input features; Inputting the node risk input features into a pre-trained node risk prediction model, extracting branches through the level features of the model, and extracting node level associated features based on the node topological relation between the belonging level labels in the node attribute vectors and the chain associated maps; And carrying out joint risk prediction on the node level association features and the node real-time risk features through a prediction output layer of the model, and outputting a node risk prediction result of a single level node comprising a node risk level, risk occurrence probability and risk type.
6. The method for managing electronic commerce data security based on big data as claimed in claim 4, wherein the step of inputting the chain association map, the risk conduction parameter and the node risk prediction result into the pre-trained chain risk conduction prediction model and outputting the link conduction risk prediction result comprises the steps of: carrying out feature fusion on node topological relation in the chain type association map, node risk conduction coefficient and side association strength in the risk conduction parameter, and node risk grade and risk occurrence probability in the node risk prediction result to obtain link risk input features; Inputting link risk input characteristics into a pre-trained chain type risk conduction prediction model, extracting branches through link topology characteristics of the model, and excavating conduction path characteristics among nodes based on a hierarchical chain type linkage relation of a chain type association map; And carrying out joint operation on the transmission path characteristics, the transmission potential energy characteristics and the risk association characteristics through a prediction output layer of the model, and outputting a link transmission risk prediction result comprising risk transmission probability, a risk transmission path, a risk influence range and a risk explosion time window.
7. The method for managing electronic commerce data security based on big data as claimed in claim 1, wherein the step of formulating the hierarchical blocking policy according to the node risk prediction result, the link conduction risk prediction result and the data sensitivity characteristic specifically comprises the steps of: Extracting and integrating risk bases of a hierarchical blocking strategy, wherein the risk bases comprise node risk levels and risk occurrence probabilities in node risk prediction results, risk conduction probabilities and risk influence ranges in link conduction risk prediction results, and sensitivity levels corresponding to data sensitivity characteristics; constructing a weighted grading model based on the risk basis, configuring differentiated weights for each risk basis, and gradually reducing weight coefficients of risk grades of corresponding nodes according to the hierarchical sequence of core hub hierarchical nodes, primary conduction hierarchical nodes and terminal associated hierarchical nodes; Weighting operation is carried out on the risk basis through the weighted hierarchical model, so that comprehensive risk values of all the level nodes and corresponding conduction links are obtained, and the comprehensive risk values are mapped into corresponding comprehensive risk levels according to a preset risk threshold; And matching the corresponding blocking strategies according to the comprehensive risk levels to obtain hierarchical blocking strategies which are matched with the risk states of the nodes of each hierarchy and the corresponding conduction links.
8. A big data based electronic commerce data security management system for implementing the steps of a big data based electronic commerce data security management method as claimed in any one of claims 1 to 7, comprising: The system comprises an acquisition module, a data processing module and a data processing module, wherein the acquisition module is used for responding to a preset response mechanism, acquiring the association relation between the node security state data of each level node of the e-commerce supply link and the nodes and preprocessing the association relation to obtain node risk characteristics, association relation characteristics and data sensitivity characteristics of each level node; The risk conduction calculation module is used for taking the node risk characteristics as the map node attributes and the association relationship characteristics as the association edge attributes to construct a chain association map, calculating the chain association map based on a preset quantitative calculation rule to obtain risk conduction parameters, wherein the risk conduction parameters comprise node risk conduction coefficients, edge risk transfer probability and edge association strength; The node risk prediction module is used for inputting node risk characteristics into a pre-trained node risk prediction model and outputting a node risk prediction result of a single node, wherein the node risk prediction result comprises a node risk level, a risk occurrence probability and a risk type; The transmission risk prediction module is used for inputting the chain type association map, the risk transmission parameters and the node risk prediction result into a pre-trained chain type risk transmission prediction model and outputting a link transmission risk prediction result, wherein the link transmission risk prediction result comprises a risk transmission probability, a risk transmission path, a risk influence range and a risk explosion time window; and the strategy making module is used for making a grading blocking strategy according to the node risk prediction result, the link conduction risk prediction result and the data sensitivity characteristic, and executing corresponding differentiated management and control on the supply link according to the grading blocking strategy.
9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of a big data based e-commerce data security management method according to any of claims 1-7 when the computer program is executed.
10. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the steps of an e-commerce data security management method based on big data as claimed in any one of claims 1 to 7.

Description

Electronic commerce data security management method and system based on big data Technical Field The invention relates to the technical field of data security management, in particular to an electronic commerce data security management method and system based on big data. Background Along with the development of digital deep upgrade and supply chain synergy of electronic commerce industry, an electronic commerce supply chain forms a core platform, namely a multi-level node linkage system of a primary supplier, a secondary server and a tertiary partner, the interaction frequency of multiple types of data such as user information, transaction data, operation data and the like among nodes is exponentially increased, the complexity and relevance of data circulation are remarkably improved, and data security becomes a core element for guaranteeing the stable operation of the electronic commerce supply chain and preventing operation risks. At present, the traditional data security management technology has obvious shortboards, key indexes such as association strength, risk conduction coefficient and the like cannot be calculated without carrying out quantitative analysis on the association relation of all links of a supply chain, so that hidden risk chain conduction paths among multiple levels of nodes cannot be explicit, key conduction paths and influence ranges cannot be identified in advance, further, risk diffusion prevention and control hysteresis is caused, meanwhile, the association risk prejudging capability is insufficient, most of passive post-event remediation after a security event is carried out, an active prejudging form combining the real-time security state and the link association characteristic of the nodes is lacking, double prejudging of the nodes and the chain conduction risk cannot be realized, a risk burst time window and an influence boundary are difficult to evaluate, a risk disposal mode is stiff, a data circulation channel is cut off in a one-tool cutting mode, the differential management and control of the risk probability and the data sensitivity level are not combined, and the risk management and control effect and the service continuity of an electronic commerce supply chain are difficult to be considered. Meanwhile, the rapid development and maturation application of the technologies such as big data, graph calculation, graph neural network, space-time sequence analysis and the like provides a feasible path for solving the technical pain point of the electronic commerce supply chain full-link data safety management and control. Based on the above, it is needed to construct an electronic commerce data security management method with a big data technology as a core support and aiming at the risk chain type conduction characteristics of the multi-level nodes of the supply chain, so as to realize full-link data association modeling, multi-dimensional risk active pre-judgment, accurate hierarchical conduction blocking and dynamic iterative optimization, break through the application limitation of the traditional technology, and improve the management and control capability and efficiency of the data security of the supply chain of the electronic commerce. Disclosure of Invention In order to overcome the defects of the prior art, the invention provides an electronic commerce data security management method and system based on big data, which are used for solving the problems in the prior art. One embodiment of the invention provides an electronic commerce data security management method based on big data, which comprises the following steps: responding to a preset response mechanism, collecting node security state data and association relation among nodes of each level of the e-commerce supply link, and preprocessing to obtain node risk characteristics, association relation characteristics and data sensitivity characteristics of the nodes of each level; calculating the chain type association map based on a preset quantitative calculation rule to obtain risk conduction parameters, wherein the risk conduction parameters comprise node risk conduction coefficients, side risk transfer probability and side association strength; Inputting node risk characteristics into a pre-trained node risk prediction model, and outputting a node risk prediction result of a single node, wherein the node risk prediction result comprises a node risk level, a risk occurrence probability and a risk type; Inputting a chain type association map, risk conduction parameters and node risk prediction results into a pre-trained chain type risk conduction prediction model, and outputting a link conduction risk prediction result, wherein the link conduction risk prediction result comprises risk conduction probability, a risk conduction path, a risk influence range and a risk burst time window; and according to the node risk prediction result, the link conduction risk prediction result and the data sensitivity characterist