T×oneHop approach for dynamic influence maximization problem

As many forms of social interactions are transient in nature, that is to say, the social network structure formed by relations between individuals evolves constantly. In practice, exploring influence maximization for product publicity and advertisement marking in dynamic networks is more challenging. Therefore, the research of dynamic influence maximization problem (DIMP) is indispensable. In this paper, we propose an efficient algorithm, T×oneHop approach, to solve the DIMP only by updating the nodes affected by the new adding seed node under the dynamic independent cascade model. It means that the seed set and pseudo-seed set always activate their neighbors within one hop in a signal snapshot graph whose total number is T. Consequently, the transfer process is similar to T×one hops of propagation in the static network. The experiment evaluations with three real social networks show that our method outperforms forward algorithm and weighted degree algorithm. In addition, our method precedes forward algorithm on running time. Finally, we find out that the common seed nodes number increases with the similarity of two consecutive snapshot graphs.