Abstract

The Internet and Social network has become the essential part of life which ease people for sharing information and media with friends. The formation of community structure is particularly important in network analysis practically. The community structure is one in which it merges the alike users by interacting with friends in social networks. Apart from this recognition of online communities users in social networks, there are various applications such as identifying a group of expert customers, a group of customers with shared endeavors, and a alike people’s group for marketing objectives. The structural Properties of social networks are analyzed with the help of Community detection algorithms. Enhanced Nearest Neighbor-Based Clustering (ENNC) is best among the community detection algorithms in which modularity based community detection is achieved in the prevailing technique. Although there exist some hindrances of overlapping communities due to the probability of user joining in more than a group is high. Hence there requires a detecting overlapping communities while analyzing realistic network. This research concentrates on Optimized Overlapping and disjoint Community Detection (OODCD) technique by adopting a by Improved Vertex Imitation Co-efficient based Community Overlap Propagation Algorithm (IVIC-COPRA).In this, the Enhanced Nearest Neighbor-Based Clustering (ENNC) approach is utilized basically which is modularity based approach for partitioning the network into minor local communities. The Louvain method is then used for detecting the disjoint community in the given network. The belonging matrix plays a vital in this research which is regularly updated whose matrix elemental value decides the role of node belonging to a community and thus the overlapping communities is found with the support of Improved Vertex Imitation Co-efficient Community Overlap Propagation Algorithm (IVIC-COPRA).The Performance of the anticipated research is examined and contrasted with the numerous eminent other overlapping community detection algorithms by means of simulation.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call