An intimacy-based algorithm for social network community detection

Abstract

Community detection is a crucial way to understand social network, and it reflects the structural characteristics of the network and the interesting features of community. We introduce the intimacy among nodes to detect community in social network. By reducing the degree of intimacy matrix between the communities, we approached the accurate community detection firstly. Then, in order to reduce the algorithm complexity, the intimacy-based algorithm for community merger is proposed. At last, compared with the existing algorithms in the theoretical and experimental respectively, we obtain that our algorithm drops the time complexity, reduces the iterations and cuts down the realization time based on the precise community detection.