LPOCSIN With K-Means: An Overlapping Clustering Technique with Cluster Information

Abstract

Research in the social network has become an escalating area. In past days, most network study on the basis of whether there is a relationship exists or not as well as study with the weight of the relationship. But the weight value was limited within the positive number. Nowadays when the social network is part of every aspect of human life, the weight value is not limited to positive numbers only. Like any online product, the user has to see the review before buying this. Today's review system of any online marketing shop is facilitated by the positive-negative value. A network with signed weight is called a signed network. It is exceptionally much vital to study with the signed network because it moreover expanding day by day. However, most of the researches in this field have done with the unweighted form of the relationship. In this paper, we will show the basic research theory of signed network like a measure of imbalance with social balance theory and clustering technique and explain how a linear programming approach works with the sign network data and giving the desired overlapping cluster.