Title: W-entropy method to measure the influence of the members from social networks

Authors: Li Weigang; Zheng Jianya; Guiqiu Liu

Addresses: Department of Computer Science, TransLab, C.P. 4466 University of Brasilia, CEP: 70910-900, Brasilia, Brazil ' Department of Computer Science, TransLab, C.P. 4466 University of Brasilia, CEP: 70910-900, Brasilia, Brazil ' School of Foreign Languages, Shenyang Normal University, Shenyang, China

Abstract: With the rapid advance of the social media, the challenge is to develop new techniques and standards to measure the influence of people or brands in the online social networks. Each website has its way of ranking the display of the most influential members of its virtual society. However, most of the current measurement methods are incomplete and one-dimensional. This paper presents a new measurement model, W-entropy, which has been developed based on information theory to study the influence of individuals based on different social networks. The model was tested using data from Facebook, Twitter, YouTube, and Google search. The proposed model can be extended to other platforms. To evaluate the effectiveness, the developed method was compared with Famecount ranking using the same data with different weight distributions. The result shows that W-entropy method is suitable for index ranking to reflect uneven information distribution across various social networks.

Keywords: entropy; metrics; information theory; social networks; W-entropy index; influential members; influence measurement; social media; Facebook; Twitter; YouTube; Google; index ranking; uneven information distribution.

DOI: 10.1504/IJWET.2013.059105

International Journal of Web Engineering and Technology, 2013 Vol.8 No.4, pp.369 - 394

Published online: 31 Mar 2014 *

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