Detecting Semantic Communities in Social Networks

Zhen LI  Zhisong PAN  Guyu HU  Guopeng LI  Xingyu ZHOU  

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E100-A   No.11   pp.2507-2512
Publication Date: 2017/11/01
Online ISSN: 1745-1337
DOI: 10.1587/transfun.E100.A.2507
Type of Manuscript: LETTER
Category: Graphs and Networks
community detection,  semantic label,  content information,  nonnegative matrix factorization,  

Full Text: PDF(1005.5KB)
>>Buy this Article

Community detection is an important task in the social network analysis field. Many detection methods have been developed; however, they provide little semantic interpretation for the discovered communities. We develop a framework based on joint matrix factorization to integrate network topology and node content information, such that the communities and their semantic labels are derived simultaneously. Moreover, to improve the detection accuracy, we attempt to make the community relationships derived from two types of information consistent. Experimental results on real-world networks show the superior performance of the proposed method and demonstrate its ability to semantically annotate communities.