Analysis of Influence Maximization in Temporal Social Networks

The problem of influence maximization aims to specify the small number of initial individuals that will eventually influence the individuals as much as possible, which has aroused wide attention of researchers. However, the most existing work is limited to the static social network and ignores the role of time in information propagation. In this paper, we analyze the influence maximization problem in temporal social networks and present a greedy-based on the latency-aware independent cascade (GLAIC) algorithm enhanced by cost-effective lazy forward optimization based on the latency-aware independent cascade model to capture the dynamic aspect of real-world social networks. Moreover, we modify the distribution of influence delays in the LAIC model by considering power-law distribution. At last, we carry out extensive experiments over the real-world networks, which demonstrate that our proposal achieves an excellent performance to other related algorithms.