Analysis of Influence Maximization Algorithm in Hub-Sparse Structure Social Network

The public opinion analysis of social network is a subject to be studied. The influence maximization is also a key issue. Research on influence maximization has been quite mature. But for the characteristics of grain security social networks, it is a new problem to find out the network's influence maximization algorithm suitable for this kind of social network. This paper analyzes the spread range and running time of three algorithms (greedy, degree centrality and PageRank) on the grain security social network through experiments. The degree centrality method takes less time to obtain similar spread range, and is more suitable for this kind of social networks. Furthermore, the concept of hub-sparse structure social network is proposed. Finally, the reason why the degree centrality method is applied to hub-sparse structure social networks is analyzed.