Title: A comprehensive analytical model of user influence in the social networks

Authors: Wang Bailing; Zhu Dongjie; Huang Junheng; Chen Zhaoqing

Addresses: School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Weihai, China ' School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Weihai, China ' School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Weihai, China ' School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Weihai, China

Abstract: The user influence analysis is one of the most interesting research topics in the field of Social Networks (SNs). The traditional methods based on degree centrality only considered the user's local influence. In recent years, the famous web ranking method (PageRank) has been widely studied in the measurement of user influence. It has achieved good results by using the global structure information of social networks. However, there are deficiencies in the contributions among users' influence. Intuitively, users' influence is closely related to the scale, solidarity and the users' status in the community. Based on the PageRank algorithm and the label propagation algorithm, this paper proposes a new comprehensive model of evaluating users' influence. This model takes into account the user's PageRank value, the user's status in the community as well as size and solidarity of the community, closer to the real situation of social networks. Experimental results show that the new model can evaluate user influence with higher efficiency and is more rational than traditional models.

Keywords: PageRank; label propagation; community; node degree; social networks.

DOI: 10.1504/IJWMC.2019.101424

International Journal of Wireless and Mobile Computing, 2019 Vol.17 No.2, pp.142 - 148

Accepted: 02 Apr 2019
Published online: 07 Aug 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article