Abstract

The identification of influential users in an online commercial social network is an important problem for companies that need to design and target more effective online marketing campaigns. In most cases, these users play roles of gate-keepers, diffusers of information, and opinion leaders who often ignite epidemics through word-of-mouth. However, since their online activities vary with time, the problem of their identification should be tackled on a temporal basis. In this paper, we model the time-dependent evolution of an online commercial social network as a time-varying weighted directed graph. Moreover, we propose an approach that takes into account behavioral and structural aspects of an online commercial social network in order to determine opinion leaders with respect to their contributions to a temporal business value of such a network.