<![CDATA[Research on the community detection in social tagging networks has attracted much attention in the last decade. Extracting the hidden topic information from tags provides a new way of thinking for community detection in social tagging networks. In this paper, a topic tagging network by extracting several topics from the tags through using the Latent Dirichlet Allocation (LDA) model is built firstly. Then a topic distance between users is defined, which depends on the bookmarking relationships between users and tags. Further, a modularity clustering approach based on the topic distance is proposed to detect communities in social tagging networks. Empirical studies on real-world networks demonstrate that the proposed method can effectively detect communities in tagging networks. DOI ; http://dx.doi.org/10.11591/telkomnika.v12i5.4170]]>
Copyrights © 2014
