A Survey on Community Mining in Social Networks

Community detection and overlapping community detection has been of significant importance in the last decade wherein invention and growth of social networks like Facebook, Twitter, Linkedin, Flickr, etc. has even made it more crucial to investigate better approaches. Overlapping community detection has many application as in modern market analysis and recommendation systems, bioinformatic systems, etc. The main hindrance in analysing social networks like Facebook is their giant social graphs and problems like detecting communities and their overlaps have very large time complexity. Finding overlapping communities in modern social graphs is an open problem and researchers have been using many heuristic graph mining and machine learning algorithms to approach the problem with less complexity. In this study we survey various community detection algorithm used in the literature as well as methods for evaluating them.