• Media type: E-Article; Electronic Conference Proceeding; Text
  • Title: Exploring Causal Relationships Among Emotional and Topical Trajectories in Political Text Data
  • Contributor: Baumann, Andreas [Author]; Hofmann, Klaus [Author]; Kern, Bettina [Author]; Marakasova, Anna [Author]; Neidhardt, Julia [Author]; Wissik, Tanja [Author]
  • imprint: Schloss Dagstuhl – Leibniz-Zentrum für Informatik, 2021
  • Language: English
  • DOI: https://doi.org/10.4230/OASIcs.LDK.2021.38
  • Keywords: topical stability ; time-series clustering ; political discourse ; Granger causality ; emotion
  • Origination:
  • Footnote: Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
  • Description: We explore relationships between dynamics of emotion (arousal and valence) and topical stability in political discourse in two diachronic corpora of Austrian German. In doing so, we assess interactions among emotional and topical dynamics related to political parties as well as interactions between two different domains of discourse: debates in the parliament and journalistic media. Methodologically, we employ unsupervised techniques, time-series clustering and Granger-causal modeling to detect potential interactions. We find that emotional and topical dynamics in the media are only rarely a reflex of dynamics in parliamentary discourse.
  • Access State: Open Access