ReaderBench: Building Comprehensive Sociograms of Online Communities A Side-by-Side Comparison between Gaming and Online Knowledge Building Communities

The usage of discussion forums and blogs has dramatically increased during the last years. Having a wide applicability, ranging from fun to learning purposes, the modeling of participants' interactions in online communities became an appealing research area. However, there are multiple factors that need to be taken into consideration when analyzing online communities, including participants' social-cognitive profiles, as well as underlying interaction patterns. The focus of this research is to introduce an integrated approach based on our ReaderBench framework together with novel comprehensive sociograms in order to analyze the interactions between participants. In addition, we perform an automated identification of underlying social-cognitive structures by applying a hierarchical clustering algorithm based on Cohesion Network Analysis (CNA) participation indices. In addition, the Actor Model from the AKKA framework was used for the parallel processing of discussions threads. In order to argue for the adequacy of the introduced mechanisms, we selected two complementary communities and we provide extensive visualizations, together with key remarks derived from the observed interaction patterns.