Automatic Recognition of Altered Argument Structures in Political Debates

Affiliation: TU Darmstadt

Associated Since: Dezember 2016

• Dr. Johannes Daxenberger, UKP Lab, TU Darmstadt
• Prof. Dr. Iryna Gurevych, UKP Lab, TU Darmstadt

CEDIFOR-Partners:

• Prof. Dr. Marcus Maurer, Institut für Publizistik, JGU Mainz
• Prof. Dr. Ivo Engels, Institut für Geschichte, TU Darmstadt

Within this project we contrastively analyze automatically recognized argumentation structures in political debates, eg. the German Parliament. Speeches as part of political debates typically involve argumentative structures such as claims. These structures vary according to context, language or speaker. The aim of this project therefore is to develop a methodology to reliably recognize argumentative structures (claims and evidences). Evaluation of these structures (over a long time span or in various thematic contexts) will be carried out in collaboration with our partners in the departments of Communication Sciences and Journalism of JGU Mainz as well as the department of History of TU Darmstadt.

Publications:

• Maurer, M., Daxenberger, J., & Gurevych, I. (2017). Argumentation Mining: Eine neue Methode zur automatisierten Textanalyse und ihre Anwendung in der Kommunikationswissenschaft. In Jahrestagung der Fachgruppe Methoden der Publizistik- und Kommunikationswissenschaft der Deutschen Gesellschaft für Publizistik- und Kommunikationswissenschaft.