Influence Minimization Algorithm Based on Coordination Game

Influence analysis is the basic technology for predicting potentially hazardous behavior and determining the traceability of the hazardous behavior in the public security domain. Previous research has focused on maximizing the diffusion of the influence; however, little research has been performed on the method of minimizing the influence of negative information dissemination in networks. This paper proposes an influence minimization algorithm based on coordinated game. When the negative information is generated in the network and some initial nodes have been infected, the goal is to minimize the number of the final infected nodes by discovering and blocking the $K$ uninfected nodes. First, the algorithm assumes that the behavior of the node propagating information depends on the coordination game with its neighboring nodes. Second, based on the local interaction model between the nodes, this paper quantifies the level of the influence of a node that is affected by its neighbors. Finally, the heuristic algorithm is used to identify the approximate optimal solution. The results of experiments performed on four real network datasets show that the proposed algorithm can suppress negative information diffusion better than the five considered existing algorithms.