Analyzing Parliamentary Questions: A Political Science Application of a New Topic Modelling Approach

Abstract

In a transdisciplinary collaboration of computer and social scientists, we have developed a novel topic discovery tool chain which - in contrast to the established probabilistic models - uses ideas from network theory and semantic word representations. The focus of this poster lies on a political science application of this new approach: analyzing written parliamentary questions. Our use case comprises foreign-policy related MIs of the last decade until April 2020 (3,700 documents). The poster presents main findings supported by diagrams derived from our network based approach.