基于数据流频繁子图挖掘的可疑金融交易动态识别

Dynamic Recognition of Suspicious Financial Transactions Based on Frequent Sub-graph Mining of Data Stream

目前,我国基于交易上报制度和静态数据挖掘的可疑金融交易识别方法存在着监测覆盖面窄、识别时效性差两大瓶颈问题。一种可行的改进是在现有方法中引入对可疑金融交易的动态识别,其中需解决的关键问题是如何及时有效地从大规模动态数据集中发现相应的可疑交易特征。设计一种基于数据流频繁子图挖掘的可疑关联特征动态识别算法,并用实验证明该算法的可行性和有效性。

At present, the limitation in real-time and coverage is the main problem that troubles the suspicious financial transaction recognition based on transactions reported system and static data mining techniques in our country. To overcome this limitation, dynamic recognition of suspicious financial transaction can be adopted. In this way, how to recognize the characteristics of suspicious transactions in large scale data set efficiently must be figured out. Hence, a new algorithm of recognizing suspicious relationship in transaction records based on frequent sub-graph mining of data stream is proposed in this paper, and it is proved to be feasible and efficient by the experiments.