Influence Maximization in Noncooperative Social Networks

In this paper, we consider the problem of maximizing information propagation with noncooperative nodes in social networks. We generalize the linear threshold model to take node noncooperation into consideration and provide a provable approximation guarantees for the noncooperative influence maximization problem. We propose an analytical model based on the generalized maximum flow problem to characterize the noncooperative behavior of an individual node in maximizing influence. Based on this, we develop a new seed node selection strategy, under the linear threshold model, to account for user noncooperativeness. Extensive simulations on large collaboration networks show that our proposed flow-based strategy outperforms the weighted degree scheme under various noncooperative scenarios. The evaluation also validates the importance of cooperation and incentives in maximizing influence.