Hits:

Indexed by:期刊论文

Date of Publication:2012-01-01

Journal:Journal of Computers

Included Journals:EI、Scopus

Volume:7

Issue:1

Page Number:135-140

ISSN No.:1796203X

Abstract:This paper introduces the concept of community seed, vector and relation matrix. In terms of the relation similarity between free vertices and the existing communities, we put vertices into different groups. A minimum similarity threshold is proposed to filter which gives a method to find the vertices located at the overlapped area between different communities. This paper analyzes a series of network dataset and proves that our algorithm is able to accurately find communities with high cohesion and weak coupling. We use a variety of test to demonstrate that our algorithm is highly effective at detecting community structures in both computer-generated and real-world networks. ? 2012 ACADEMY PUBLISHER.