Community Detection in Large-Scale Social Networks Using Various Similarity Indexes and Genetic Algorithm

Abstract

These days, community detection is identified as an important research topics in online social networks. Communities exist in social networks are the important feature which is deliberated as a prospective stricture in forming the performance of the social individuals. In social network Analysis lot of researchers got attracted towards the Community detection problem. It is considered as one of the main interesting issues with high complexity in complex structure processing. This issue can also be minimized to different set of small problem of graph theory like triangle counting, clique identification, etc. It is also evident that in large scale social networks, this problem is NP-complete problem. In past literature most of the algorithms proposed on meta-heuristic ways are having modularity maximization as an objective function and aim is to optimize this modularity value. In many cases, it is notices that identified communities are largely dependent on the edges present in the network and may suffer from resolution limit. To resolve this issue, in this paper we calculated the similarity index values among each pair of nodes is measured in a parallel manner and an objective function based on the similarity index is proposed and processed through a Genetic algorithm. The proposed algorithm is related with traditional algorithms and tested on various real-world datasets