Analyzing internet topics by visualizing microblog retweeting

摘要

Microblog is a large-scale information sharing platform where retweeting plays an important role in information diffusion. Analyzing retweeting evolutions can help reasoning about the trend of public opinions. Information visualization techniques are used to demonstrate the retweeting behavior in order to understand how Internet topics diffuse on Microblogs. First, a graph clustering method is used to analyze the retweeting relationships among people of different occupations. Then a new algorithm based on electric field is proposed to visualize the layout of the relationship links. A prediction method based on three diffusion models is presented to predict the number of retweets over time. Finally, three real world case studies show the validity of our methods. (C) 2014 Elsevier Ltd. All rights reserved.