MEI: Mutual Enhanced Infinite Community-Topic Model for Analyzing Text-Augmented Social Networks

Community and topic can help summarize the text-augmented social networks. Existing works mixed up community and topic by regarding them as the same. However, there is inherent difference between community and topic such that treating them as the same is not so flexible. We propose a mutual enhanced infinite community-topic model (MEI) to detect communities and topics simultaneously in text-augmented social networks. Community and topic are correlated via community-topic distribution. The mutual enhancement effect between community and topic is validated by introducing two novel measures perplexity with community (perplexityc) and MRK with topic (MRKt). To determine the numbers of communities and topics automatically, Dirichlet Process Mixture model (DPM) and Hierarchical Dirichlet Process mixture model (HDP) are used to model community and topic respectively. We further introduce parameter to model the weight of community and topic responsible for community inference. Experiments on the co-author network built from a subset of DBLP data show MEI outperforms the baseline models in terms of generalization performance. Parameter study shows that MEI is averagely improved by 3.7% and 15.5% in perplexityc and MRKt respectively by setting low weight for topic. We also experimentally validate that MEI can determine the appropriate numbers of communities and topics.