Containment of Misinformation Propagation in Online Social Networks with Given Deadline.

Nan Wang,Li Yu,Ni Ding,Dong Yang
2014-01-01
Abstract:In recent years, online social networks have become one of the most influential mediums of sharing information. However, alongside this promising feature comes the threat of rumors and misinformation propagation, which has aroused extensive attention from our society. In this paper, we try to figure out how to select the smallest set of highly influential nodes decontaminated with good information to contain the propagation of misinformation effectively within given deadline and reach the ratio of positive nodes ωand the growth rate of newly added nodes with misinformation φrequired in this paper. In the social networks with time delay, two competitive strategies of containing competitive propagating model (IC-CCPM) are proposed, respectively First Come, First Served strategy and Simultaneous Competitive strategy, considering not only Timeliness of competition between good information and misinformation, but also two kinds of information’s arrival at the same time, when users’ choices depend on the influence probability of neighbor users. Two containing strategies are developed to select the smallest K positive seed nodes, respectively Greedy Containing strategy and Boundary Containing strategy. A lot of experiments are made to test our proposed strategies by using three datasets, respectively twitter, Friendster and random. The experiment results show that Greedy Containing strategy and Boundary Containing strategy can be applied to most of the distributions of initial set of nodes contaminated by misinformation, which are superior to other available methods. In addition, boundary containing strategy is similar with greedy containing strategy containing effect, but better than greedy containing strategy in time efficiency.
What problem does this paper attempt to address?