Detection of Multi-Relations Based on Semantic Communities Behaviors

In this paper, we represent a novel approach for detecting multi-relations among network community related activities. By using semantic information and social network analysis (SNA). We can mining various features of groups of people through their network text inter-communications, such as E-mail or BBS and the like. Our method also differentiates and labels distinct relations between people from different aspects of their works, which may help security or management officers to be objective through analyzing the framework of the semantic communities distribution, to extract the clues they need to support the early preventions or controls. The dataset used to validate this study is a collection of Email data from one of Chinese research unit, and the E-mail records from Enron corp., which is retrieved freely from the Internet. This method could be applied to many similar researches for some given purposes.