PROJECT TITLE :

Overlapping Community Detection Using Neighborhood-Inflated Seed Expansion

ABSTRACT:

Community detection is a vital task in network analysis. A community (additionally called a cluster) is a set of cohesive vertices that have more connections inside the set than outside. In many social and information networks, these communities naturally overlap. For example, in an exceedingly social network, each vertex in a very graph corresponds to a personal who typically participates in multiple communities. During this paper, we tend to propose an efficient overlapping community detection algorithm using a seed enlargement approach. The key idea of our algorithm is to search out smart seeds, and then greedily expand these seeds based on a community metric. Inside this seed expansion method, we have a tendency to investigate the matter of how to see smart seed nodes in a graph. In explicit, we have a tendency to develop new seeding ways for a personalised PageRank clustering theme that optimizes the conductance community score. An important step in our technique is that the neighborhood inflation step where seeds are modified to represent their entire vertex neighborhood. Experimental results show that our seed growth algorithm outperforms different state-of-the-art overlapping community detection methods in terms of manufacturing cohesive clusters and identifying ground-truth communities. We conjointly show that our new seeding strategies are higher than existing ways, and are thus effective to find good overlapping communities in real-world networks.