基于交易网络特征向量中心度量的可疑洗钱识别系统

Suspicious money laundering detection system based on eigenvector centrality measure of transaction network

针对可疑洗钱识别问题,提出了基于交易网络特征向量中心度量的可疑洗钱行为检测系统。检测系统包含交易数据预处理、特征向量中心度量、时序检测分析三个关键子算法。同时还给出了三种用于检测异常交易行为的度量指标。通过对银行交易数据进行的实验仿真,验证了系统的有效性。

For anti-money laundering, a detection system based on eigenvector centrality measure was introduced, which included pre-processing of transaction data, eigenvector centrality measure and time-series analysis. Three key indexes of suspicious activities detection were also provided. Through the simulation on the transaction data of bank, the validity of the detection system was proved.