Mining Communities and Their Relationships in Blogs: A Study of Online Hate Groups

This article outlines a semi-automated approach for analyzing the content and structure of online hate groups active on blogging platforms. The authors used a social network analysis (SNA) approach that consisted of four modules: blog spider, information extraction, network analysis, and visualization. This approach was then applied to identify and analyze a selected set of 28 anti-black hate groups (820 bloggers in total), on Xanga, one of the most popular blog hosting websites. The network analysis methods included topological analysis, centrality analysis, community analysis, including blockmodeling, and visualization, with emphases on multidimensional scaling, graph layout algorithms, and other features. Based on their methodology and examination of extremist groups, the authors were able to map network and community structures and locate centers of influence. The analysis results revealed interesting demographical and topological characteristics present within these groups, and effectively identified two large communities along with a number of smaller communities.

This article will be of particular interest to researchers and practitioners interested in exploring methods that can identify and assess online extremist communities, and especially those active within blogging websites. The study demonstrated the feasibility of applying a semi-automated approach to the study of hate groups and other related communities in blogs. The approach and methodology employed could be useful to the mapping of various types of extremist networks, identifying community structures, and locating centers of influence across a broad range of extremist group typologies.