Influence Blocking Maximization in Social Network Using Centrality Measures

Online social networks play an important role as a suitable platform for information diffusion. While positive news diffusion on social network has a great impact in people's life, the negative news can also spread as fast as positive ones. To make the social network a reliable place, it is necessary to block inappropriate, unwanted and contamination diffusion. In this paper, we study the notion of competing negative and positive campaigns in a social network and address the influence blocking maximization (IBM) problem to minimize the bad effect of misinformation. IBM problem can be summarized as identifying a subset of nodes to adopt the positive influence under Multi-campaign Independent Cascade Model (MCICM) as diffusion model to minimize the number of nodes that adopt the negative influence at the end of both propagation processes. We proposed Centrality_IBM algorithm based on centrality measures for finding an appropriate candidate subset of nodes for spreading positive diffusion in order to minimizing the IBM problem. Then, we experimentally compare the performance of the proposed algorithm using some centrality measures to choose the appropriate subset of positive influential nodes. The experiments on different real datasets reveal that the closeness centrality measure outperforms the alternative centrality measures in most of the cases.