Optimal Control of the Adversarial Information Propagation in Online Social Networks

The popularization of online social networks (OSNs) opens opportunities for information sharing. The existing works mainly investigate the propagation of the single type of information. However, the propagation of adversarial information, that is, the positive and negative information, still remains largely unexplored to date. In this paper, a propagation model is proposed to characterize the propagation dynamics of the adversarial information in OSNs. Afterwards, two different control measures are designed to control the dynamic propagation process and a new control model with control measures is obtained. Then, the minimization of the total system costs is modeled as an optimal control problem. Finally, we obtain an optimal control system using Pontryagin maximum principle. Experiments on the real-world data set show that our proposed control measures is efficient in controlling the propagation of the adversarial information.