Design And Implementation Of Financial Transaction Money Laundering Detection System

With the rampant upstream crimes such as telecom fraud,tax evasion,illegal financing,drug trafficking,corruption and bribery,China's money laundering activities have become increasingly large.The scale and amount of money laundering cases are expanding year by year,posing a serious threat to China's economic system and national security.At present,the police use manual operation to detect money laundering cases,which is extremely inefficient and prone to errors.Therefore,how to efficiently import the information involved in the case,analyze and judge the case,track the flow of funds,dig up suspected money laundering gangs,provide investigators with quick investigation ideas and improve the efficiency of case handling has become an urgent problem to be solved.The research background of this system is the detection of money laundering cases by XX police.Financial transactions are taken as the main research object,which mainly includes six modules: data import and persistence,data operation,account correlation analysis,suspicious gang topology diagram,case management and system management.B/S architecture is adopted,front-end jQuery is responsible for topology and data display,and back-end SSM framework is responsible for business logic and data access.Aiming at the problem of different data formats and precision used in the transaction flow of major banks,the "self-identification" and "manual import" methods are designed to realize the efficient and fast import system of transaction flow;realize the functions of data search,statistical analysis and map display to facilitate the analysis of information involved in the case;the most obvious feature of money laundering is the existence of a loop;starting from "suspicious path tracing" and "associated account tracing",a money laundering loop model is established to trace the money laundering path of card number and account.;it is difficult for analysts to mine the whole money laundering gang simply by manually selecting Excel tables,therefore,a node intimacy calculation method is proposed to define the transaction relationship between accounts,and an improved Louvain-Method is used to build a suspicious gang identification model to quickly mine money laundering gangs.Finally,the "force-oriented visualization" layout is realized,including account node,account relationship,drag and drop,layout adjustment,"fixed stillness" or "force-bearing activity" to provide visual interactive support.Finally,the system test can efficiently and accurately import the involved information into the system;excavate suspicious circuits and provide money laundering clues;analystsclearly acquire the organizational structure of money laundering groups.At present,the system meets the needs of public security analysts in the daily detection of money laundering cases,and has assisted public security in the detection of many money laundering cases.