Negative influence minimizing by blocking nodes in social networks

Social networks are becoming vital platforms for the spread of positive information such as innovations and negative information propagation like malicious rumors. In this paper, we address the problem of minimizing the influence of negative information. When negative information such as a rumor emerges in the social network and part of users have already adopted it, our goal is to minimize the size of ultimately contaminated users by discovering and blocking k uninfected users. A greedy method for efficiently finding a good approximate solution to this problem is proposed. The comparison experimental results on the Enron email network dataset demonstrate our proposed method is more effective than centrality based methods, such as degree centrality, betweenness centrality and PageRank.