Abstract

Virtual human communities that exist on the Internet reflect social relationships between people. They can be either directly maintained by computer systems like Frindster, LinkedIn or extracted from data about user activities in the communication network like e-mails, chats, blogs, homepages connected by hyperlinks, etc. Social network analysis can be applied to virtual communities and can deliver interesting knowledge about particular network members as well as cohesive subgroups. It appears that there is a great need to find important individuals or a set of people who would represent larger communities. These people would be able to perform specific tasks or could become a target group either for marketing or advertising purposes. The new research on representative discovery for human communities based on social network analysis is presented in this paper. Its main goal is to improve the process of target group selection by adding the social elements derived from the behaviours of people. The target group should fulfil a set of prerequisites. Simultaneously, the selection process ought to respect specific social aspects provided by the social network. The entire selection process considered in the paper is called human filtering and includes one of two separate scenarios: selection by single member or selection by social group.