A Novel Method of Identifying Influential Users on Social Network
Jingchi Jiang, Wenchong Bi, Chengqi Yi, Yuanyuan Bao, Yibo Xue · 2015
Identifying the influential users on social network is a critical problem for advertising, promotion, public opinion analysis and network security.A few empirical studies have verified that some users could lead to wider spreading and promote the information diffusion prominently.However, the existing studies limited by a single layer analysis for information diffusion, there is still lack of a multifaceted analysis method for identifying influential nodes.In this paper, we propose a novel method for identifying the influential users in the topology of information diffusion, which contains a three-tier analytical mechanism: information-level analysis, relationship-level analysis and community-level analysis.This three-tier analytical mechanism can upgrade information-level to relationship-level, and finally identifying the influential nodes in relationship-level.The experiments on Sina Weibo demonstrate that 0.3% of users who span structural holes control 28% of the information diffusion in each event and can cause the secondary-wave of information diffusion.When the information diffuses to users with higher pagerank, more users will participate in the information diffusion.