基于社会网络分析法的洗钱犯罪数据挖掘侦查技术的改进
Improvement of Investigation Techniques in Money Laundering Crime's Data Mining Based on SNA
牛惊雷 1牛易航2
作者信息
- 1. 中国人民警察大学,河北 廊坊 065000
- 2. 天津市公安局河西区公安分局,天津 300202
- 折叠
摘要
随着互联网信息技术的发展,洗钱犯罪手段、方式愈发复杂和隐蔽,极大地影响了可疑交易识别精度,使侦查工作陷入低效的困境,进一步改进洗钱犯罪数据挖掘技术是必然趋势."结构二重性"理论拓展侦查视角,社会网络分析法揭示了可疑交易的识别不仅需要结合大量账户信息和交易特征,更需要深度考量与账户交易关联的网络结构特征.图卷积神经网络理论作为能够深度融合基础特征与社会网络结构特征的学习算法,为洗钱犯罪数据挖掘侦查技术提供了改进途径.
Abstract
The development of Internet information technology has made the criminal means and methods of money laundering crime increasingly complex and concealed,therefore substantially impairing the accuracy of suspicious transaction identification and plunging the investigation work into an inefficient dilemma.Therefore,the enhancement of the data mining techniques for the detection of money laundering crime has become an inexorable trend.The"duality of structure"theory helps expand the perspective of investigation,while social network analysis reveals that the identification of suspicious transactions requires not only extensive account information and transaction features,but also an in-depth examination of the network structure associated with account transactions.The theory of Graph Convolution Network,or GCN,as a learning algorithm capable of deeply integrating basic features with social network structure attributes,provides an approach for improving investigation techniques applied in money laundering crime's data mining.
关键词
洗钱犯罪/可疑交易/社会网络/图卷积神经网络Key words
money laundering crime/suspicious transactions/social networks/Graph Convolution Network引用本文复制引用
出版年
2024
NETL
NSTL
维普
万方数据