overlapping community detection in social networks based on stochastic simulation

community detection is a task of fundamental importance in social network analysis. community structures enable us to discover the hidden interactions among the network entities and summarize the network information that can be applied in many applied domains such as bioinformatics, finance, e-commerce and forensic science. there exist a variety of methods for community detection based on different metrics and domain of applications. most of these methods are based on the existing of the non-overlapping or sparse overlapping communities. moreover, the experimental analysis showed that, overlapping areas of communities become denser than non-overlapping area of communities. in this paper, significant methods of overlapping community detection are compared according to well-known evaluation criteria. the experimental analyses on artificial network generation have shown that earlier methods of community detection will not discover overlapping communities properly and we offered suggestions for resolving them.