A semantic overlapping community detection algorithm based on field sampling

The traditional semantic social network (SSN) community detection algorithms need to preset the number of the communities and could not detect the overlapping communities. To solve the issue of presetting the number of communities, we present a clustering algorithm for community detection based on the link-field-topic (LFT) model suggested. For the process of clustering is independent of context sampling, the number of communities is not necessary to be preset. To solve the issue of overlapping community detection, we establish the semantic link weight (SLW) depending on the analysis of LFT, to evaluate the semantic weight of links for each sampling field. The proposed clustering algorithm is based on the SLW which could separate the SSN into clustering units. As a result, the intersection on several units is the overlapping part. Finally, we establish semantic modularity (SQ) involving SQ1 and SQ2 for the evaluation of the detected semantic communities. The efficiency and feasibility of the LFT model and the semantic modularity is verified by experimental analysis.