A method and system for constructing a financial anti-fraud business database

By constructing a financial transaction relationship diagram and analyzing transaction data, identifying key nodes and abnormal fluctuations, and monitoring and simulating the fund and credit status in real time, the problem of insufficient real-time and preventiveness of traditional financial anti-fraud databases is solved, and complex finance is achieved In-depth understanding and accurate identification of trading models have improved the risk management capabilities and market adaptability of financial institutions.

CN118916348BActive Publication Date: 2024-12-03SICHUAN FENLING INFORMATION TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411414730.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-10-11
Publication Date
2024-12-03
Estimated Expiration
2044-10-11

AI Technical Summary

Technical Problem

Traditional financial anti-fraud business databases have shortcomings in real-time and preventiveness, making it difficult to accurately identify and track fraudulent activities in complex financial transaction models, and lack the ability to respond quickly to market changes, resulting in poor monitoring and preventive measures.

Method used

By constructing a financial transaction relationship chart, analyzing transaction time, amount and participant identity, identifying key nodes and abnormal fluctuations, monitoring transaction activities in real time, simulating changes in funds and credit status, generating financial network analysis data, conducting data classification and risk assessment, Build a financial anti-fraud database.

Benefits of technology

It improves the accuracy and real-time response capabilities of data analysis, can accurately identify risks and fraud, reduce economic losses caused by fraud, and enhances the risk management capabilities and market adaptability of financial institutions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118916348B_ABST
    Figure CN118916348B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of financial information technology, specifically to a method and system for constructing a financial anti-fraud business database, comprising the following steps: based on transaction business record information, analyzing transaction time, transaction amount and participant identity information, constructing a financial transaction relationship diagram, and generating financial transaction network information. In the present invention, by analyzing financial transaction records, constructing financial transaction network information, increasing the depth of data processing, and improving the accuracy of data analysis through network construction, by analyzing the statistical volatility of transaction activities, accurately identifying abnormal fluctuation nodes, identifying risks and fraudulent behaviors, real-time monitoring and simulation of node funds and credit status changes, real-time response to market changes, taking measures before risks occur, reducing economic losses caused by fraud, combining classification and risk assessment of transaction data, improving financial institutions' risk management capabilities and market adaptability, and enhancing customer trust and market competitiveness.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of financial information technology, and in particular to a method and system for constructing a financial anti-fraud business database. Background Art

[0002] The field of financial information technology focuses on applying a variety of information technologies to improve the efficiency, security, and innovation of financial services. It combines computing technology, network security, big data analysis, and artificial intelligence technologies to optimize financial processes, enhance customer service, and strengthen risk management. Financial information technology plays a core role in a variety of financial institutions, including banks, insurance companies, and investment companies. Through automated processing and data analysis, it helps institutions improve transaction efficiency, strengthen compliance monitoring, and reduce financial fraud and various types of financial crimes through risk assessment.

[0003] Among them, the construction method of the financial anti-fraud business database focuses on building a database system for detecting and avoiding financial fraud, including multiple processes of data collection, cleaning, integration, and analysis to achieve the identification, monitoring, and prevention of fraudulent behavior. By integrating and analyzing a variety of financial transaction data, potential fraudulent behavior can be identified. By utilizing data analysis technology and machine learning algorithms, financial activities can be monitored in real time to detect abnormal behavior at an early stage, helping financial institutions reduce economic losses caused by fraud, enhance the risk management capabilities of financial institutions, and improve market credibility and customer trust.

[0004] The construction method of traditional financial anti-fraud business database still has certain limitations in terms of real-time and preventiveness. The ability to monitor and prevent financial fraud in real time is insufficient, resulting in time lags in identifying and responding to potential risks, increasing the economic losses of financial institutions when facing fraud. In terms of data integration and analysis, it relies on fixed models and algorithms, which limits the in-depth understanding and analysis of complex financial transaction patterns, making it difficult to accurately identify and track financial fraud activities conducted through complex transaction networks, and difficult to flexibly adapt to the rapid changes in the financial market. It lacks the ability to quickly respond to and accurately predict emerging transaction patterns, resulting in poor results of monitoring and prevention measures. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method and system for constructing a financial anti-fraud business database.

[0006] In order to achieve the above object, the present invention adopts the following technical solution, a method for constructing a financial anti-fraud business database, comprising the following steps:

[0007] S1: Based on the transaction business record information, analyze the transaction time, transaction amount and participant identity information, build a financial transaction relationship diagram, and generate financial transaction network information;

[0008] S2: Based on the financial transaction network information, analyze the connection status of multiple nodes, identify key nodes in the flow of funds, and evaluate the transmission efficiency of multiple nodes in the financial network to generate key node analysis results;

[0009] S3: According to the key node analysis results, collect transaction data associated with multiple nodes, calculate the statistical volatility of transaction activities of multiple nodes, identify abnormal fluctuation nodes, and obtain node risk assessment data;

[0010] S4: Based on the node risk assessment data, by real-time monitoring of the transaction activities of abnormal fluctuation nodes, recording transaction frequency, amount changes and transaction connections, identifying abnormal transaction behaviors, and generating abnormal transaction monitoring records;

[0011] S5: Based on the abnormal transaction monitoring records, by simulating the changes in the capital and credit status of multiple nodes, the dynamic changes in the capital flow and credit flow in the transaction network are calculated to generate financial network analysis data;

[0012] S6: Based on the financial network analysis data, the transaction data is classified, regular and risky transactions are recorded and a data set is constructed, the consistency and integrity of the data is verified, and a financial anti-fraud database is generated.

[0013] As a further solution of the present invention, the financial transaction network information includes trader identity identification, transaction relationship topology structure, and transaction amount statistical data; the key node analysis results include the node's betweenness centrality score, the node's closeness centrality score, and the node's degree centrality score; the node risk assessment data includes the node's transaction frequency fluctuation data, the node's transaction amount fluctuation information, and the node's transaction type diversity index; the abnormal transaction monitoring records include abnormal transaction frequency, abnormal transaction amount change information, and abnormal transaction patterns; the financial network analysis data includes capital flow concentration analysis results, credit risk aggregation area analysis data, and dynamic stability indicators of the transaction network; the financial anti-fraud database includes a regular transaction data set, a risky transaction data set, and a transaction data information entropy value.

[0014] As a further solution of the present invention, based on the transaction business record information, the transaction time, transaction amount and participant identity information are analyzed to construct a financial transaction relationship diagram, and the steps of generating financial transaction network information are specifically as follows:

[0015] S101: Based on the transaction business record information, analyze and record the transaction time, amount and participant identity information to generate a basic transaction data set;

[0016] S102: Based on the basic transaction data set, a directed graph of financial activities is constructed by mapping transaction participants as nodes, transactions as edges, and transaction amounts as edge weights to generate a financial transaction graph;

[0017] S103: According to the financial transaction graph, by analyzing the connection density and transaction volume of the nodes, the weight of the edge is adjusted to reflect the importance and frequency of the transaction, the structure of the network graph is optimized, and the financial transaction network information is generated.

[0018] As a further solution of the present invention, based on the financial transaction network information, the connection status of multiple nodes is analyzed, key nodes in the capital flow are identified, and the transmission efficiency of multiple nodes in the financial network is evaluated. The steps of generating the key node analysis results are specifically as follows:

[0019] S201: Based on the financial transaction network information, analyzing the connection status of multiple nodes in the graph, calculating the characteristic vector centrality according to the number of interactions between the multiple nodes, evaluating the activity of the nodes in the network, and generating node connection status data;

[0020] S202: Based on the node connection status data, calculate the transmission efficiency of multiple nodes, evaluate the speed and breadth of capital flow, analyze the average path length and transaction efficiency between nodes, and generate node transmission efficiency data;

[0021] S203: According to the node transmission efficiency data, by analyzing the capital flow and transaction frequency of multiple nodes, key nodes are identified and recorded, and key node analysis results are generated.

[0022] As a further solution of the present invention, the formula for calculating the eigenvector centrality is specifically:

[0023]

[0024] in, For Node The eigenvector centrality of the node is used to measure the The influence in the network reflects the node The centrality accumulated through its network connections, is the maximum eigenvalue of the adjacency matrix, which is used for normalized centrality calculation to ensure the stability of the results and the effectiveness of comparison. is the total number of nodes in the network, ensuring that the analysis covers all nodes in the entire network. is the node in the adjacency matrix and nodes The elements between are used to indicate whether there is a direct connection between two nodes and to evaluate whether there is a direct connection between nodes. For Node The eigenvector centrality of the node For Node The contribution of the centrality is It is the node index in the network, used to traverse the nodes All adjacent nodes of The centrality is considered when all Directly connected nodes, It is the node index of the current analysis, used to identify the target node in the network, distinguish multiple nodes in the network, ensure that the centrality of each node can be calculated independently and consider its relative importance in the network, calculate the influence of multiple nodes in the financial network through the formula, conduct financial network analysis of node influence, identify key nodes and conduct risk assessment.

[0025] As a further solution of the present invention, according to the key node analysis results, the transaction data associated with multiple nodes are collected, the statistical volatility of the transaction activities of multiple nodes is calculated, the abnormal fluctuation nodes are identified, and the steps of obtaining the node risk assessment data are specifically as follows:

[0026] S301: According to the key node analysis result, collect transaction data associated with the key node, record the transaction time and transaction amount of the target node, and generate a transaction activity data set;

[0027] S302: According to the transaction activity data set, the transaction volatility indexes of multiple nodes are evaluated through time series analysis to obtain node volatility analysis data;

[0028] S303: Based on the node volatility analysis data, analyze and identify nodes whose volatility exceeds a set threshold, record transaction information of the target node, and generate node risk assessment data.

[0029] As a further solution of the present invention, based on the node risk assessment data, by real-time monitoring of the transaction activities of abnormally fluctuating nodes, recording transaction frequency, amount changes and transaction connections, identifying abnormal transaction behaviors, and generating abnormal transaction monitoring records, the specific steps are:

[0030] S401: Based on the node risk assessment data, real-time monitoring and recording of transaction activities of abnormally fluctuating nodes, recording the transaction time, amount and counterparty information, and generating a real-time transaction monitoring log;

[0031] S402: Analyze the real-time transaction monitoring log, compare the current transaction mode of the abnormal node with the normal mode, analyze and identify the transaction behavior that is inconsistent with the normal mode, and generate an abnormal transaction record;

[0032] S403: Based on the abnormal transaction records, by analyzing the pattern deviation behavior and evaluating the impact on the financial network, the amount, frequency and transaction connection information of the abnormal behavior are recorded to generate an abnormal transaction monitoring record.

[0033] As a further solution of the present invention, according to the abnormal transaction monitoring records, by simulating the changes in the capital and credit status of multiple nodes, the dynamic changes of the capital flow and credit flow in the transaction network are calculated, and the steps of generating financial network analysis data are specifically as follows:

[0034] S501: According to the abnormal transaction monitoring record, by collecting the capital and credit flow information of the node, simulating the change of capital and credit status, recording the daily capital and credit changes of multiple nodes, and generating capital credit dynamic data;

[0035] S502: Analyze the capital credit dynamic data, calculate the change rate and concentration trend of capital flow and credit flow of multiple nodes, identify the key paths of capital flow and credit flow, and generate flow analysis data;

[0036] S503: Based on the flow analysis data, by evaluating the capital inflow and outflow ratios, credit change frequency and amplitude of multiple nodes and regions, identifying abnormal capital flow concentration and credit risk nodes, and generating financial network analysis data.

[0037] As a further solution of the present invention, based on the financial network analysis data, the transaction data is classified, conventional and risky transactions are recorded and a data set is constructed, and the consistency and integrity of the data are verified to generate a financial anti-fraud database. Specifically, the steps are as follows:

[0038] S601: Based on the financial network analysis data, by identifying and classifying regular transactions and risky transactions, organizing the data structure and constructing a transaction data set, generating a transaction data classification record;

[0039] S602: Based on the transaction data classification records, by performing uncertainty analysis on the transaction data, evaluating the randomness level in the data set, and generating randomness evaluation data;

[0040] S603: Based on the randomness evaluation data, the classified data is subjected to consistency check and integrity verification to optimize the accuracy and integrity of the data and generate a financial anti-fraud database.

[0041] A system for constructing a financial anti-fraud business database, the system for constructing a financial anti-fraud business database is used to execute the above-mentioned method for constructing a financial anti-fraud business database, the system comprising:

[0042] The transaction network construction module builds a financial transaction relationship diagram based on transaction business record information by analyzing transaction time, amount and participant identity. It also identifies and evaluates key nodes in the flow of funds by evaluating the connection status of multiple nodes and generates transaction node analysis data.

[0043] The transaction data analysis module calculates the volatility of transaction activities of multiple nodes based on the transaction node analysis data, identifies nodes with abnormal volatility, and obtains node volatility assessment results;

[0044] The abnormal monitoring module monitors abnormal fluctuation nodes in real time according to the node volatility assessment results, records transaction frequency, amount changes and transaction connections, identifies abnormal transaction behaviors, and generates abnormal node transaction records;

[0045] The network risk analysis module uses the abnormal node transaction records to simulate the changes in funds and credit status, analyze the dynamic changes in the fund flow and credit flow in the transaction network, identify the nodes where funds flow to, and generate risk node analysis data;

[0046] The database construction module analyzes the data according to the risk nodes, constructs data sets of regular and risky transactions by classifying the transaction data, verifies the integrity of the data, and generates a financial anti-fraud database.

[0047] Compared with the prior art, the advantages and positive effects of the present invention are:

[0048] In the present invention, by analyzing financial transaction records, constructing financial transaction network information, increasing the depth of data processing, and improving the accuracy of data analysis through network construction, by analyzing the statistical volatility of transaction activities, accurately identifying abnormal fluctuation nodes, identifying risks and fraudulent behaviors, real-time monitoring and simulation of node capital and credit status changes, real-time response to market changes, taking measures before risks occur, reducing economic losses caused by fraud, combining classification and risk assessment of transaction data, improving financial institutions' risk management capabilities and market adaptability, and enhancing customer trust and market competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the workflow of the present invention;

[0050] Figure 2 This is a detailed flow chart of S1 of the present invention;

[0051] Figure 3 This is a detailed flow chart of S2 of the present invention;

[0052] Figure 4 This is a detailed flow chart of S3 of the present invention;

[0053] Figure 5This is a detailed flow chart of S4 of the present invention;

[0054] Figure 6 This is a detailed flow chart of S5 of the present invention;

[0055] Figure 7 This is a detailed flow chart of S6 of the present invention;

[0056] Figure 8 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0058] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0059] See also Figure 1 The present invention provides a technical solution, a method for constructing a financial anti-fraud business database, comprising the following steps:

[0060] S1: Based on the transaction business record information, analyze the transaction time, transaction amount and participant identity information, build a financial transaction relationship diagram, and generate financial transaction network information;

[0061] S2: Based on the financial transaction network information, analyze the connection status of multiple nodes, identify the key nodes in the capital flow, and evaluate the transmission efficiency of multiple nodes in the financial network to generate key node analysis results;

[0062] S3: Based on the key node analysis results, collect transaction data associated with multiple nodes, calculate the statistical volatility of transaction activities of multiple nodes, identify abnormal fluctuation nodes, and obtain node risk assessment data;

[0063] S4: Based on the node risk assessment data, by real-time monitoring of the transaction activities of abnormally fluctuating nodes, recording the transaction frequency, amount changes and transaction connections, identifying abnormal transaction behaviors, and generating abnormal transaction monitoring records;

[0064] S5: Based on abnormal transaction monitoring records, by simulating the changes in the capital and credit status of multiple nodes, the dynamic changes in the capital flow and credit flow in the transaction network are calculated to generate financial network analysis data;

[0065] S6: Based on financial network analysis data, classify transaction data, record routine and risky transactions and build data sets, verify data consistency and integrity, and generate a financial anti-fraud database.

[0066] Financial transaction network information includes trader identity, transaction relationship topology, and transaction amount statistics. Key node analysis results include the node's betweenness centrality score, the node's closeness centrality score, and the node's degree centrality score. Node risk assessment data includes the node's transaction frequency fluctuation data, the node's transaction amount fluctuation information, and the node's transaction type diversity index. Abnormal transaction monitoring records include abnormal transaction frequency, abnormal transaction amount change information, and abnormal transaction patterns. Financial network analysis data includes capital flow concentration analysis results, credit risk concentration area analysis data, and dynamic stability indicators of the transaction network. The financial anti-fraud database includes regular transaction data sets, risky transaction data sets, and transaction data information entropy values.

[0067] See also Figure 2 Based on the transaction business record information, the transaction time, transaction amount and participant identity information are analyzed to construct a financial transaction relationship diagram. The specific steps for generating financial transaction network information are as follows:

[0068] S101: Based on the transaction business record information, analyze and record the transaction time, amount and participant identity information to generate a basic transaction data set;

[0069] In sub-step S101, based on the transaction business record information, the transaction time, transaction amount, and identity information of the transaction participants are collected and recorded in real time through the transaction system database, and the data is sorted and classified and archived in the basic transaction data set. The data set will serve as the basic platform for subsequent analysis. By performing data integrity verification on each transaction record, the accuracy and completeness of the data are ensured. The data set serves as the data basis for financial institutions to conduct analysis, supporting various financial analysis activities such as transaction pattern identification, risk assessment, and customer behavior analysis, and is used to enhance the accuracy and efficiency of data-driven decision-making.

[0070] S102: Based on the basic transaction data set, a directed graph of financial activities is constructed by mapping transaction participants as nodes, transactions as edges, and transaction amounts as edge weights to generate a financial transaction graph;

[0071] In sub-step S102, based on the basic transaction data set, a graph theory representation of nodes and edges is used to map transaction participants and transaction behaviors, where each transaction participant is regarded as a node, and each transaction is weighted by the edge connecting the nodes according to the amount. By identifying and understanding the financial network structure, the main paths and risk nodes of capital flow are analyzed, and the generated financial transaction graph provides an intuitive network analysis tool to support the operation of network analysis algorithms, including path analysis, centrality measurement and community detection.

[0072] S103: According to the financial transaction graph, by analyzing the connection density and transaction volume of the nodes, the weight of the edge is adjusted to reflect the importance and frequency of the transaction, the structure of the network graph is optimized, and the financial transaction network information is generated;

[0073] In the above content, based on the financial transaction graph, by analyzing the connection density and transaction volume of the nodes, according to the formula , calculate the new weight of each edge, Is a node and The new weights of the edges between them reflect the importance and frequency of transactions, is the transaction amount, and The nodes are and The degree of connectivity is used to map the actual situation of financial activities and generate financial transaction network information.

[0074] In the formula, represents the adjusted edge weight, Represents the transaction amount between nodes, and Representative Node and degree of connectivity.

[0075] Detailed explanation of the formula and calculation process:

[0076] Assume Node and The transaction amount between 10,000 yuan, node Connectivity is 8, nodes Connectivity If it is 12, the calculation is as follows:

[0077]

[0078]

[0079]

[0080] The results show that, considering the node and After the total number of connections, the transaction arrive The relative weight is adjusted to 500 to evaluate the actual importance of multiple nodes in the financial network.

[0081] See also Figure 3 , based on the financial transaction network information, the connection status of multiple nodes is analyzed, the key nodes in the capital flow are identified, and the transmission efficiency of multiple nodes in the financial network is evaluated. The steps to generate the key node analysis results are as follows:

[0082] S201: Based on the financial transaction network information, the connection status of multiple nodes in the graph is analyzed, and according to the number of interactions between the multiple nodes, the characteristic vector centrality is calculated, the activity of the nodes in the network is evaluated, and the node connection status data is generated;

[0083] The formula for calculating the eigenvector centrality is:

[0084]

[0085] in, For Node The eigenvector centrality of the node is used to measure the The influence in the network reflects the node The centrality accumulated through its network connections, is the maximum eigenvalue of the adjacency matrix, which is used for normalized centrality calculation to ensure the stability of the results and the effectiveness of comparison. is the total number of nodes in the network, ensuring that the analysis covers all nodes in the entire network. is the node in the adjacency matrix and nodes The elements between are used to indicate whether there is a direct connection between two nodes and to evaluate whether there is a direct connection between nodes. For Node The eigenvector centrality of the node For Node The contribution of the centrality is It is the node index in the network, used to traverse the nodes All adjacent nodes of The centrality is considered when all Directly connected nodes, It is the node index of the current analysis, used to identify the target node in the network, distinguish multiple nodes in the network, ensure that the centrality of each node can be calculated independently and consider its relative importance in the network, calculate the influence of multiple nodes in the financial network through the formula, conduct financial network analysis of node influence, identify key nodes and conduct risk assessment.

[0086] formula:

[0087]

[0088] Detailed explanation of the formula and the process of formula calculation and derivation:

[0089] Formula used to calculate the node The eigenvector centrality of , measure the influence of nodes in financial networks, analyze the relative importance of each node in the entire network structure by expressing the centrality of each node as the weighted sum of the centralities of directly connected nodes, and identify key nodes in the network;

[0090] Parameter meaning and setting value:

[0091] For Node The eigenvector centrality is used to measure the relative influence of a node in the entire network.

[0092] is the maximum eigenvalue of the adjacency matrix, which is used for normalization to ensure that the calculation result of the centrality will not be biased by the size of the eigenvalue. It is assumed to be 2.5, reflecting the maximum connection strength of the network.

[0093] is the total number of nodes in the network, assuming it is 100, which means the analyzed network includes 100 nodes.

[0094] is the node in the adjacency matrix and nodes The elements between indicate whether there is a direct connection between the two nodes. 1 indicates that there is a connection, and 0 indicates that there is no connection. Is a node The eigenvector centrality of the node For Node Contribution of centrality, assuming that node There are three directly connected neighbors ,in, , , , , , ;

[0095] Substitute the parameters into the formula for calculation:

[0096]

[0097]

[0098]

[0099] result Indicates the node The centrality score in the network is 0.4, which is relatively low, indicating that it has little influence in the financial transaction network, but it is still an active participant in the network. The value is used to evaluate and monitor the activity and importance of multiple nodes in the network.

[0100] S202: Based on the node connection status data, the transmission efficiency of multiple nodes is calculated, and the node transmission efficiency data is generated by evaluating the speed and breadth of capital flow, analyzing the average path length and transaction efficiency between nodes;

[0101] In sub-step S202, based on the node connection status data, the transmission efficiency of multiple nodes is calculated. By analyzing the speed and breadth of capital flow between nodes, the average path length between nodes and capital flow data are used to measure the transaction efficiency of the entire network. By analyzing the direct and indirect connections between each node, the average path length between nodes is calculated, the circulation speed of funds and information in the network is evaluated, and the efficient nodes and bottleneck nodes in the network are identified to optimize transaction paths and enhance overall network performance, generate node transmission efficiency data, and optimize the response time and transaction processing capabilities of financial services.

[0102] S203: According to the node transmission efficiency data, by analyzing the capital flow and transaction frequency of multiple nodes, identifying and recording key nodes, and generating key node analysis results;

[0103] In sub-step S203, the capital flow and transaction frequency of each node in the network are analyzed based on the node transmission efficiency data. By analyzing the capital flow and transaction data of each node, the key nodes that play a core role in the network are identified, including analyzing the node's capital receipt and payment situation, the frequency and scale of transactions, the connection density and strategic position in the network. The key nodes are identified by analyzing the key capital flow and information flow, and the key node analysis results are generated to provide risk assessment and decision support for financial institutions, monitor and manage key risk points in the network.

[0104] See also Figure 4,According to the key node analysis results, the transaction data associated with multiple nodes are collected, the statistical volatility of transaction activities of multiple nodes is calculated, and the abnormal fluctuation nodes are identified. The specific steps to obtain the node risk assessment data are as follows:

[0105] S301: According to the key node analysis results, collect transaction data associated with the key nodes, record the transaction time and transaction amount of the target node, and generate a transaction activity data set;

[0106] In sub-step S301, based on the key node analysis results, transaction data directly associated with the key nodes are collected. By analyzing the transaction time and transaction amount of the target node, the financial transaction database is used to identify the transaction records associated with the key nodes, including transaction timestamps and amounts, and build a transaction activity data set for the key nodes. The data set includes the time point and amount of each transaction, which is used to provide basic data for pattern analysis and risk assessment, track and monitor the financial activities of the key nodes, and identify abnormal transaction behaviors to achieve risk management and compliance monitoring.

[0107] S302: According to the transaction activity data set, through time series analysis, the transaction volatility indicators of multiple nodes are evaluated to obtain node volatility analysis data;

[0108] In sub-step S302, based on the transaction activity data set, the transaction volatility of each key node is evaluated through time series analysis technology, including analyzing the time series data of the transaction amount of each node, calculating statistical indicators, including mean, standard deviation, kurtosis and skewness, and identifying transaction volatility patterns. The volatility indicator is used to reveal the stability of the capital flow of each node and the trend of changes in multiple time periods. The node volatility analysis data provides a quantitative perspective to evaluate the financial stability of each node and predict potential risks, and formulate targeted risk mitigation measures.

[0109] S303: Based on the node volatility analysis data, analyze and identify nodes whose volatility exceeds a set threshold, record the transaction information of the target node, and generate node risk assessment data;

[0110] In the above content, the nodes whose volatility exceeds the set threshold are analyzed and identified, according to the formula Assess the transaction risk of each node and generate node risk assessment data.

[0111] In the formula, represents the risk assessment value, is the standard deviation of the node transaction amount, reflecting the volatility of the transaction. It is the risk sensitivity coefficient, which is set based on historical data and market conditions.

[0112] Detailed explanation of the formula and calculation process:

[0113] Assume that the standard deviation of the transaction amount of a node is , risk sensitivity coefficient Set it to 1.5, and calculate the risk assessment value of the node as:

[0114]

[0115] The results show that the risk assessment value of the node is 3000, and the node is classified as a risky node. It is necessary to pay attention to transaction behavior and capital flow to prevent risk events.

[0116] See also Figure 5 Based on the node risk assessment data, by real-time monitoring of the transaction activities of abnormally fluctuating nodes, recording the transaction frequency, amount changes and transaction connections, identifying abnormal transaction behaviors, the specific steps for generating abnormal transaction monitoring records are as follows:

[0117] S401: Based on the node risk assessment data, real-time monitoring and recording of transaction activities of abnormally fluctuating nodes, recording the transaction time, amount and counterparty information, and generating a real-time transaction monitoring log;

[0118] In sub-step S401, based on the node risk assessment data, the transaction activities of the nodes identified as abnormal fluctuations are monitored and recorded in real time, including the time of occurrence of the transaction, the transaction amount and the information of the other party. By capturing the transaction behavior of each abnormal node and analyzing the deviation from the historical transaction pattern, including the payment method, source and flow of funds, the generated real-time transaction monitoring log provides real-time and accurate transaction data support for security audits and risk management, and enhances the responsiveness and anti-fraud effect of the monitoring system.

[0119] S402: Analyze the real-time transaction monitoring log, compare the current transaction mode of the abnormal node with the normal mode, analyze and identify the transaction behavior that is inconsistent with the normal mode, and generate abnormal transaction records;

[0120] In sub-step S402, the real-time transaction monitoring log is used to analyze the current transaction pattern of each abnormal node, and by comparing it with the regular transaction pattern in the history of the node, including comparison of multiple dimensions such as transaction frequency, amount, transaction counterparty and transaction time, through cluster analysis and outlier detection, transaction behaviors that deviate from the regular pattern, including fraudulent behavior or risk events, are identified, and abnormal transaction records are generated, which provide a decision-making basis for the risk management department of the financial institution to investigate financial risks and take risk mitigation measures to ensure the stability of the financial system and the safety of customer assets.

[0121] S403: Based on the abnormal transaction records, by analyzing the pattern deviation behavior and evaluating the impact on the financial network, the amount, frequency and transaction connection information of the abnormal behavior are recorded to generate abnormal transaction monitoring records;

[0122] In the above content, by analyzing the pattern deviation behavior and evaluating its impact on the financial network, according to the formula Calculate the risk index for each abnormal trading behavior , generate abnormal transaction monitoring records.

[0123] In the formula, A risk index representing abnormal trading behavior, is the amount of abnormal transaction, is the frequency of abnormal transactions, is the length of the transaction connection, indicating the number of nodes involved.

[0124] Detailed explanation of the formula and calculation process:

[0125] Assume an abnormal transaction with a transaction amount of The frequency of occurrence is times, the node connection length of the transaction is nodes, the risk index is calculated as:

[0126]

[0127] The results show that the risk index of abnormal transactions is 6666.67, indicating that it poses a major risk to the financial network and requires timely response and processing.

[0128] See also Figure 6 According to the abnormal transaction monitoring records, by simulating the changes in the capital and credit status of multiple nodes, the dynamic changes of capital flow and credit flow in the transaction network are calculated, and the steps for generating financial network analysis data are as follows:

[0129] S501: Based on the abnormal transaction monitoring records, by collecting the capital and credit flow information of the nodes, simulating the changes in capital and credit status, recording the daily capital and credit changes of multiple nodes, and generating capital credit dynamic data;

[0130] In sub-step S501, based on abnormal transaction monitoring records, by collecting fund and credit flow information related to abnormal nodes, the changes in fund and credit status of each node are simulated, including recording the inflow and outflow of funds and changes in credit status of multiple nodes in daily operations. The data is integrated and analyzed through simulation technology to construct fund credit dynamic data. The data set provides financial institutions with real-time financial health monitoring, allowing financial institutions to instantly understand the fund and credit status of each node, prevent financial crises, and use it for risk management and decision support.

[0131] S502: Analyze the dynamic data of capital credit, calculate the change rate and concentration trend of capital flow and credit flow of multiple nodes, identify the key paths of capital flow and credit flow, and generate flow analysis data;

[0132] In sub-step S502, based on the dynamic data of funds and credit, the key paths in the financial network are identified by analyzing the change rate and concentration trend of the funds flow and credit flow of each node, including calculating the funds inflow and outflow rate of each node, the frequency and amplitude of credit changes, and the concentration of funds and credit distribution in the network, and generating flow analysis data. The data is used to reveal the core paths of funds and credit flows in the financial network.

[0133] S503: Based on the flow analysis data, by evaluating the capital inflow and outflow ratios, credit change frequency and amplitude of multiple nodes and regions, identifying abnormal capital flow concentration and credit risk nodes, and generating financial network analysis data;

[0134] In the above content, by evaluating the capital inflow and outflow ratio, credit change frequency and amplitude of multiple nodes and regions, according to the formula Assess capital flows and credit risk, and generate financial network analysis data.

[0135] In the formula, Represents the overall risk index of the financial network, is the change in capital flow, is the fund flow risk weight, is the credit change, is the credit risk weight, is the number of nodes evaluated.

[0136] Detailed explanation of the formula and calculation process:

[0137] Assume that in the target area, the change in capital flow Yuan, fund flow risk weight , credit change , credit risk weight , the number of evaluation nodes , substitute the parameters into the formula for calculation:

[0138]

[0139] The results showed that the overall financial network risk index of the target area was 93, indicating that the area needed increased monitoring and risk management measures.

[0140] See also Figure 7, based on financial network analysis data, classify transaction data, record routine and risky transactions and build data sets, verify data consistency and integrity, and generate financial anti-fraud databases in the following steps:

[0141] S601: Based on the financial network analysis data, by identifying and classifying regular transactions and risky transactions, organizing the data structure and constructing the transaction data set, generating transaction data classification records;

[0142] In sub-step S601, based on the financial network analysis data, the transaction activities in the network are analyzed to identify and classify them into regular transactions and risky transactions. By evaluating the characteristics of various transactions, including the size of the transaction amount, the transaction frequency, the credit history of the participants, and the geographical location where the transaction occurs, the data structure is organized and a transaction data set including two major categories of transactions is constructed. The generated transaction data classification records distinguish between regular transactions and risky transactions, and provide basic data for risk management and monitoring activities, which are used to strengthen the identification and prevention of potential risky transactions.

[0143] S602: Based on the classified records of transaction data, by performing uncertainty analysis on the transaction data, the randomness level in the data set is evaluated to generate randomness evaluation data;

[0144] In sub-step S602, based on the transaction data classification records, uncertainty analysis of the transaction data is performed to evaluate the randomness level in the data set. Statistical methods, including coefficient of variation and entropy analysis, are used to measure the uncertainty and variability in the data, helping financial institutions to parse the random distribution in the data set, evaluate data quality issues and identify atypical patterns in the data. The generated randomness assessment data provides an important reference for financial institutions for data processing and analysis, ensuring the reliability of the data set and the validity of the analysis results.

[0145] S603: Based on the random evaluation data, the classified data is checked for consistency and integrity, the accuracy and integrity of the data are optimized, and a financial anti-fraud database is generated;

[0146] In the above content, the classified data is checked for consistency and integrity, according to the formula Optimize data accuracy and completeness and generate a financial anti-fraud database.

[0147] In the formula, represents the data quality index, is the data consistency check score, is the data integrity verification score, is the total number of data items.

[0148] Detailed explanation of the formula and calculation process:

[0149] Assume that the consistency check score in the dataset , integrity verification score , total number of data items , substitute the parameters into the formula for calculation:

[0150]

[0151] The results show that the data quality index is 0.925, indicating that the data quality of the financial anti-fraud database is very high, providing financial institutions with a reliable and accurate data source to support complex risk analysis and decision making.

[0152] See also Figure 8 A system for constructing a financial anti-fraud business database, the system for constructing a financial anti-fraud business database is used to execute the above-mentioned method for constructing a financial anti-fraud business database, and the system includes:

[0153] The transaction network construction module builds a financial transaction relationship diagram based on transaction business record information by analyzing transaction time, amount and participant identity. It also identifies and evaluates key nodes in the flow of funds by evaluating the connection status of multiple nodes and generates transaction node analysis data.

[0154] The transaction data analysis module analyzes data based on transaction nodes, calculates the volatility of transaction activities of multiple nodes, identifies nodes with abnormal volatility, and obtains node volatility assessment results;

[0155] The abnormal monitoring module monitors abnormal fluctuation nodes in real time based on the node volatility assessment results, records transaction frequency, amount changes and transaction connections, identifies abnormal transaction behaviors, and generates abnormal node transaction records;

[0156] The network risk analysis module uses abnormal node transaction records to simulate changes in capital and credit status, analyze the dynamic changes in capital flow and credit flow in the transaction network, identify capital flow concentration nodes, and generate risk node analysis data;

[0157] The database construction module analyzes data according to risk nodes, classifies transaction data, constructs data sets for regular and risky transactions, verifies the integrity of the data, and generates a financial anti-fraud database.

[0158] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for constructing a financial anti-fraud business database. It is characterized in that The following steps are involved: Based on the transaction business record information, the transaction time, transaction amount and participant identity information are analyzed to build a financial transaction relationship diagram and generate financial transaction network information, including: Based on the transaction business record information, analyze and record the transaction time, amount and participant identity information to generate a basic transaction data set; Based on the basic transaction data set, a directed graph of financial activities is constructed by mapping transaction participants as nodes, transactions as edges, and transaction amounts as edge weights to generate a financial transaction graph; According to the financial transaction graph, by analyzing the connection density and transaction volume of the nodes, adjusting the weight of the edge to reflect the importance and frequency of the transaction, optimizing the structure of the network graph, and generating financial transaction network information; Based on the financial transaction network information, the connection status of multiple nodes is analyzed, key nodes in the capital flow are identified, and the transmission efficiency of multiple nodes in the financial network is evaluated to generate key node analysis results, including: Based on the financial transaction network information, the connection status of multiple nodes in the graph is analyzed, and according to the number of interactions between the multiple nodes, the characteristic vector centrality is calculated to evaluate the activity of the nodes in the network and generate node connection status data; Based on the node connection status data, the transmission efficiency of multiple nodes is calculated, and the node transmission efficiency data is generated by evaluating the speed and breadth of capital flow, analyzing the average path length and transaction efficiency between nodes; According to the node transmission efficiency data, by analyzing the capital flow and transaction frequency of multiple nodes, key nodes are identified and recorded, and key node analysis results are generated; According to the key node analysis results, the transaction data associated with multiple nodes are collected, the statistical volatility of the transaction activities of multiple nodes is calculated, the abnormal fluctuation nodes are identified, and the node risk assessment data is obtained; Based on the node risk assessment data, by real-time monitoring of the transaction activities of abnormally fluctuating nodes, recording transaction frequency, amount changes and transaction connections, identifying abnormal transaction behaviors, and generating abnormal transaction monitoring records; According to the abnormal transaction monitoring records, by simulating the changes in the capital and credit status of multiple nodes, the dynamic changes in the capital flow and credit flow in the transaction network are calculated to generate financial network analysis data; Based on the financial network analysis data, the transaction data is classified, routine and risky transactions are recorded and a data set is constructed, the consistency and integrity of the data is verified, and a financial anti-fraud database is generated.

2. The method for constructing a financial anti-fraud service database according to claim 1, It is characterized in that The financial transaction network information includes the trader identity, the transaction relationship topology, and the transaction amount statistics. The key node analysis results include the node's betweenness centrality score, the node's closeness centrality score, and the node's degree centrality score. The node risk assessment data includes the node's transaction frequency fluctuation data, the node's transaction amount fluctuation information, and the node's transaction type diversity index. The abnormal transaction monitoring records include abnormal transaction frequency, abnormal transaction amount change information, and abnormal transaction patterns. The financial network analysis data includes the capital flow concentration analysis results, credit risk aggregation area analysis data, and the dynamic stability index of the transaction network. The financial anti-fraud database includes a regular transaction data set, a risky transaction data set, and a transaction data information entropy value.

3. The method for constructing a financial anti-fraud service database according to claim 1, It is characterized in that The formula for calculating the eigenvector centrality is specifically: , in, For Node The eigenvector centrality of the node is used to measure the The influence in the network reflects the node The centrality accumulated through its network connections, is the maximum eigenvalue of the adjacency matrix, which is used for normalized centrality calculation to ensure the stability of the results and the effectiveness of comparison. is the total number of nodes in the network, ensuring that the analysis covers all nodes in the entire network. is the node in the adjacency matrix and nodes The elements between are used to indicate whether there is a direct connection between two nodes and to evaluate whether there is a direct connection between nodes. For Node The eigenvector centrality of the node For Node The contribution of the centrality is It is the node index in the network, used to traverse the nodes All adjacent nodes of The centrality is considered when all Directly connected nodes, It is the node index of the current analysis, used to identify the target node in the network, distinguish multiple nodes in the network, ensure that the centrality of each node can be calculated independently and consider its relative importance in the network, calculate the influence of multiple nodes in the financial network through the formula, conduct financial network analysis of node influence, identify key nodes and conduct risk assessment.

4. The method for constructing a financial anti-fraud service database according to claim 1, It is characterized in that According to the key node analysis results, the transaction data associated with multiple nodes are collected, the statistical volatility of the transaction activities of multiple nodes is calculated, the abnormal volatility nodes are identified, and the steps of obtaining the node risk assessment data are specifically as follows: According to the key node analysis results, the transaction data associated with the key nodes is collected, the transaction time and transaction amount of the target nodes are recorded, and a transaction activity data set is generated; According to the transaction activity data set, the transaction volatility indicators of multiple nodes are evaluated through time series analysis to obtain node volatility analysis data; Based on the node volatility analysis data, nodes whose volatility exceeds a set threshold are analyzed and identified, transaction information of the target nodes is recorded, and node risk assessment data is generated.

5. The method for constructing a financial anti-fraud service database according to claim 1, It is characterized in that Based on the node risk assessment data, by real-time monitoring of the transaction activities of abnormal fluctuation nodes, recording transaction frequency, amount changes and transaction connections, identifying abnormal transaction behaviors, and generating abnormal transaction monitoring records, the specific steps are: Based on the node risk assessment data, real-time monitoring and recording of transaction activities of abnormally fluctuating nodes, recording the transaction time, amount and counterparty information, and generating a real-time transaction monitoring log; Analyze the real-time transaction monitoring log, compare the current transaction mode of the abnormal node with the normal mode, analyze and identify the transaction behavior that is inconsistent with the normal mode, and generate abnormal transaction records; According to the abnormal transaction records, by analyzing the pattern deviation behavior and evaluating the impact on the financial network, the amount, frequency and transaction connection information of the abnormal behavior are recorded to generate abnormal transaction monitoring records.

6. The method for constructing a financial anti-fraud service database according to claim 1, It is characterized in that According to the abnormal transaction monitoring records, by simulating the changes in the capital and credit status of multiple nodes, the dynamic changes in the capital flow and credit flow in the transaction network are calculated, and the steps of generating financial network analysis data are specifically as follows: According to the abnormal transaction monitoring records, by collecting the capital and credit flow information of the nodes, simulating the changes in capital and credit status, recording the daily capital and credit changes of multiple nodes, and generating capital credit dynamic data; Analyze the capital credit dynamic data, calculate the change rate and concentration trend of capital flow and credit flow of multiple nodes, identify the key paths of capital flow and credit flow, and generate flow analysis data; Based on the flow analysis data, by evaluating the capital inflow and outflow ratios, credit change frequency and amplitude of multiple nodes and regions, abnormal capital flow concentrations and credit risk nodes are identified to generate financial network analysis data.

7. The method for constructing a financial anti-fraud service database according to claim 1, It is characterized in that Based on the financial network analysis data, the transaction data is classified, routine and risky transactions are recorded and a data set is constructed, and the consistency and integrity of the data are verified. The specific steps of generating a financial anti-fraud database are as follows: Based on the financial network analysis data, by identifying and classifying regular transactions and risky transactions, organizing data structures and constructing transaction data sets, and generating transaction data classification records; Based on the classified records of the transaction data, by performing uncertainty analysis on the transaction data, the randomness level in the data set is evaluated to generate randomness evaluation data; Based on the random evaluation data, the classified data is checked for consistency and verified for integrity, the accuracy and integrity of the data are optimized, and a financial anti-fraud database is generated.

8. A system for building a financial anti-fraud business database. It is characterized in that According to the method for constructing a financial anti-fraud service database according to any one of claims 1 to 7, the system comprises: The transaction network construction module builds a financial transaction relationship diagram based on transaction business record information by analyzing transaction time, amount and participant identity. It also identifies and evaluates key nodes in the flow of funds by evaluating the connection status of multiple nodes and generates transaction node analysis data. The transaction data analysis module calculates the volatility of transaction activities of multiple nodes based on the transaction node analysis data, identifies nodes with abnormal volatility, and obtains node volatility assessment results; The abnormal monitoring module monitors abnormal fluctuation nodes in real time according to the node volatility assessment results, records transaction frequency, amount changes and transaction connections, identifies abnormal transaction behaviors, and generates abnormal node transaction records; The network risk analysis module uses the abnormal node transaction records to simulate the changes in funds and credit status, analyze the dynamic changes in the fund flow and credit flow in the transaction network, identify the nodes where funds flow to, and generate risk node analysis data; The database construction module analyzes the data according to the risk nodes, constructs data sets of regular and risky transactions by classifying the transaction data, verifies the integrity of the data, and generates a financial anti-fraud database.