Signed Integrated PageRank for Rapid Information Diffusion in Online Social Networks

The social networking platforms have become interactive online interfaces to capture human behavioral dynamics by analyzing the information spreading phenomena across the network. Moreover, the pattern of information diffusion is largely governed by human emotions, which can be modeled suitably through online signed social networks. Our research considers two friendship sociograms modeled as signed social networks having positive and negative relations. Our main objective is to minimize the number of initial set of influencers, thereby forming a minimum set of triggering nodes to obtain maximum diffusion in the social network. This would further assist in the decision making of identifying a minimal set of nodes to trigger maximum information spread in the network. An integrated page ranking algorithm SIPR-k± is proposed to find the number of such initial nodes. We also experimentally proved that in case of both the social networks, only 3.12–4.37% triggering nodes were required by our proposed algorithm to achieve maximum diffusion. Comparative studies were performed with different versions of the existing models to justify enhanced performance of our approach.