Political debates between legislators and political organizations are traditionally analyzed using qualitative content analysis of text data. In recent years, however, the relational dimension of political discourse was rediscovered. Attempts were made at formalizing content analysis by drawing on social network analysis and related approaches. In previous research, new cross-sectional and longitudinal models based on social network analysis were devised for analyzing policy debates. A software package called Discourse Network Analyzer (DNA) (and additional bindings for the statistical programming environment R) was developed to implement these new methods. This research project aims at (1) extending the current state of the methodology, (2) implementing new state-of-the-art methodological innovations from computer science and physics into the Discourse Network Analyzer, and (3) applying these innovations to real-world political discourses in order to understand this apparently ill-defined phenomenon of political discourse in a much better way.
