A HEURISTIC APPROACH FOR INFERRING LINKS IN SOCIAL NETWORK

Abstract

Social Networking Sites are experiencing a rapid growth; there seems to be no limit to their size. Many Social Networking Sites boast with millions of members using their networks on regular basis to communicate, share, create, and collaborate with others. Size of Social networks are changing over time because of addition and deletion of nodes and edges. Given a snapshot of a social network at some moment of time, the link prediction problem is the problem of inferring the new links of the network at some later moment. Existing methods for inferring the links is based solely on the information based on the network topology.. In this paper we suggest a method that uses some extra information besides that given in the network topology. We assume that such information (represented as a knowledge dependency graph) may be obtained from an expert system (or human expertise). We consider the co-authorship domain, and illustrate our method using this domain.