Influence maximization based on activity degree in mobile social networks

The problem of influence maximization (IM) has become an important research topic due to the rapid growth of mobile social networks. It attempts to identify a set of nodes, referred to as influencers, contributing to the spread of maximum information. In this article, we present the construction of social relation graph based on mobile communication data. And we propose a new centrality measure-activity degree to characterize the activity of nodes. By combining the local attributes of nodes and the behavioral characteristics of nodes to measure node activity degree, which can be used to evaluate the influence of users in mobile social networks, we introduce Susceptible-Infected-Susceptible model to simulate the dynamic spreading of information. We take advantage of the two indicators the degree centrality and the betweenness centrality to get a better ranking results. In comparison with spanning graph and initial graph, the results of comparison demonstrate that our algorithm has advantages in the scope of influence propagation.