Influence Maximization Across Multiple Social Networks

Influence maximization has always been a research hotspot in the field of social networks.In the past,research was often conducted on the basis of a single social networking platform.However,with the rapid development of social networks,various social networking platforms with different focuses have emerged.Online social life of users often involves multiple social networking platforms at the same time.Therefore,only considering the maximization of the influence of a single social network platform can not achieve a real maximum impact.This paper studies the issue of influence maximization across multiple social networks and proposes an effective solution.The main content of this paper can be summarized as follows.First,this paper proposes a multi-social-network information diffusion model,and then defines the influence maximization problem of multiple social networks based on this model.In order to reduce the complexity of the problem,this paper proposes a network aggregation method that aggregates multiple social networks into a single social network from a macro perspective,thereby transforming the influence maximization problem of multiple social networks into the influence maximization problem of a single social network.Experiments show the effectiveness of the aggregation method in the transformation of the problem.Second,in order to solve the influence maximization problem of a single social network,this paper proposes a community-based method for influence maximization.The method includes two parts:a community detection algorithm and an influence maximization algorithm.This paper first proposes a local community detection method to divide the network into communities that meet the needs of community-based influence maximization research.Subsequently,this paper proposes a method to maximize the influence,which takes advantages of features of communities to simplify the influence maximization process of the greedy algorithm.Experiments show that our community detection algorithm is more suitable for community-based influence maximization research,and our influence maximization algorithm has better performance than the comparison algorithms.