Exponential Random Graph Modeling for Micro-blog Network Analysis

Social network analysis is used to study complex networks by analyzing static structure and dynamic changes. Nowadays micro-blog as a new social media is becoming the most popular communication platform. How to capture micro-blog network structure especially dynamic structure poses more scientific interest. In this paper, we choose Chinese micro-blog, Sina weibo, on topic of diabetes as our test bed. We calculate degree, average shortest path, betweenness and clustering coefficient to analyze its static structure. More important works, we introduce a general model for micro-blog with directed network data, Exponential-family Random Graph Models (ERGMs), and illustrate the utility for modeling, analyzing and simulating micro-blog network. We also provide the goodness-of-fit approach to capture and reproduce the structure of the fitted micro-blog network. We demonstrate the characteristic results of average degree, diameter and clustering coefficient of diabetes microblog static structure. Parameters estimation of model, similarity results of simulated networks and observed networks, and goodness of fit analysis for micro-blog network are all illustrated that ERGMs are excellent methods to deeply capture the complex network structure.