Analysis of User Roles and the Emergence of Themes in Discussion Forums

This work explores network analysis methods for the analysis of emergent themes as well as types of users in discussion forums. The paper provides both, a description of the analysis approach and its application as a case study. To that end, keywords are extracted from forum threads and then linked to the forum users resulting in a bipartite network based on their activity in discussion threads. Applying bipartite clustering on those networks, groups of users with common interest in themes can be identified as well as groups of related keywords based on their common relations to users. As a case study, the approach is applied to a discussion forum of a Courser a MOOC. The results reveal some interesting patterns and phenomena of thematic development that take place in such large scale learning courses.