Topic modeling of German Parliament Speeches with Regard to Sustainability Aspects

Topic modeling of German Parliament Speeches with Regard to Sustainability Aspects
Thomas Schmidt
Angela Ruf
Christian Wolff
Digital Humanities, Topic Modeling, Parliamentary Analysis, NLP


Discussion about sustainability has become more intense over the last years and manifested a huge debate in politics and society. Although it seems to be more likely related to buzz words like climate change and renewable energy, sustainable development can be divided into social, economical and ecological aspects. The setup of Agenda 2030 (Sustainable Development Goals) in 2015 by the United Nations General Assembly with 17 interlinked global goals, reaffirms the political importance of this subject.

Zielsetzung der Arbeit

The aim of this study is to examine how political debate on sustainability has developed in the German parliament between 1996 and 2021. Furthermore we want to evaluate how topics in this debate have changed over six legislative periods and if different parties represent varying roles. Therefore a structured corpus of speeches related to sustainability issues has to be created and used for further analysis.

Konkrete Aufgaben

  • Related work
  • Construction of a corpus including all plenary protocols in the German parliament between 1996 and 2021
  • Development of several parsers that identify speeches and additional metadata like party, speaker, year and legislative period to save the results in a database
  • Query and filter out those speeches related to sustainability
  • Explorative corpus analysis like n-grams, word frequencies and diachronic analysis
  • Further data processing to perform topic modeling
  • Report and visualization of the results

Erwartete Vorkenntnisse

  • Python and Natural Language Processing (NLP)
  • Corpus preparation of large amounts of textual data

Weiterführende Quellen

Blätte, A. and Blessing, A. (2018). The GermaParl Corpus of Parliamentary Protocols. In Proceedings of the International Conference on Language Resources and Evaluation, LREC 2018, 7-12 May 2018, Miyazaki, Japan.

Müller-Hansen, F., Callaghan, M. W., Lee, Y. T., Leipprand, A., Flachsland, C., Minx, J. C. (2021): Who cares about coal? Analyzing 70 years of German parliamentary debates on coal with dynamic topic modeling. - Energy Research and Social Science, 72, 101869.

Walter, T., Kirschner, C., Eger, S., Glavaš, G., Lauscher, A., & Ponzetto, S. P. (2021). Diachronic analysis of german parliamentary proceedings: Ideological shifts through the lens of political biases. In 2021 ACM/IEEE Joint Conference on Digital Libraries (JCDL) (pp. 51-60). IEEE.