Static Analysis and Exponential Random Graph Modelling for Micro-Blog Network
Dong-Hui Yang,Guang Yu
DOI: https://doi.org/10.1177/0165551513512251
2013-01-01
Journal of Information Science
Abstract:Social network analysis has been used to study complex networks by analysing their static structure and the dynamic changes. Although one of the newer forms of social media, micro-blogs have quickly become one of the most popular communication platforms. This popularity accounts for, in part, an increase in the scientific interest in micro-blogs and their users. In this paper, we chose as our test bed diabetes-related posts from the Chinese micro-blog Sina Weibo. We calculated the degree, average shortest path, betweenness and clustering coefficient of the Sian Weibo network to analyse its static structure. We demonstrate the characteristic results of average degree, diameter and clustering coefficient of diabetes micro-blog static structure. More importantly, we introduce a general model for micro-blog with directed network data, Exponential-family Random Graph Models (ERGMs). Meanwhile, we illustrate the utility for estimating, analysing and simulating micro-blog network. We also provide a goodness-of-fit approach to capture and reproduce the structure of the fitted micro-blog network. Parameter estimation of the model, similarity results of simulated networks and observed networks, and goodness of fit analysis for the micro-blog network all illustrate that ERGMs are excellent methods for deeply capturing complex network structures.