基于有向赋权图的垃圾邮件社团发现算法

Mail Community Detecting algorithm based on Weighted and Directed Graph

本文在有向赋权邮件网络的基础上,针对垃圾邮件的特征,分析邮件发送者和接收者的关系,通过社团密度这一度量函数,对有向赋权邮件网络进行社团结构分析,以准确确定社团结构与形式,该算法另一优势是可发现重叠社团。

In this paper, we analysis the feature of spam and the relationship between email sender and receiver, propose a dimension function of community density which based on weighted and directed email community. Then analysis the structure of weighted and directed email network, determine the community structure and form. To find the overlapping communities is another advantage of this algorithm.