Discovering Topic-Oriented Highly Interactive Online Communities.

摘要: Community detection is an interesting field of online social networks. Most existing approaches either consider common attributes network users or rely on only connections among the users. However, not enough attention paid to degree interactions community members in retrieved communities, resulting less interactive members. This inactivity will create problems for many businesses as they require highly efficiently advertise their marketing information. In this paper, we propose a model detect topic-oriented densely-connected communities which have active each other. We conduct experiments real dataset demonstrate effectiveness our proposed approach.