A Core Theory Based Algorithm for Influence Maximization in Social Networks

Abstract

The connectivity of large scale complex networks relies on a specific small set of structural nodes which is called the core of the whole network. The influence maximization problem is to identify such set of nodes, known as influencers, who can trigger the maximum range of information propagation in a network, which is one of the most important problems in network science. In this paper, we introduce core theory and simulated annealing algorithm to locate the set of core nodes. The initial active influencer can be acquired by optimally choosing from the core nodes. We compare our method with other alternative algorithms in real-world datasets. The results demonstrate that our method is competitive in both information propagation efficiency and time-consuming in all the diffusion models we consider.