Influence maximization based on the realistic independent cascade model

In order to propagate information through the social network, how to find a seed set that can affect the maximum number of users is named as influence maximization problem. A lot of works have been done on this problem, mainly including two aspects: establishing a reasonable information diffusion model and putting forward the appropriate seeding strategy. However, there are few models in the existing ones that consider the acceptance probability of candidate seed nodes in social networks. So in this paper, we consider and solve this problem by introducing a more realistic model, which is the proposed Realistic Independent Cascade (RIC) model. Based on the RIC model, many state-of-the-art seeding algorithms perform not so well because there is no mechanism on dealing with the acceptance probability. So based on the RIC model, we propose a new seeding strategy which is called R-greedy. Furthermore, M-greedy algorithm is proposed to reduce the time complexity of R-greedy. Then, D-greedy algorithm which not only increased the performance but also reduced the time complexity of R-greedy is proposed by combining the advantages of R-greedy and M-greedy. Experiments on the real-world networks and synthetic networks demonstrate that the proposed R-greedy, M-greedy and D-greedy algorithms outperforms state-of-the-art algorithms.