New corpus analyzes German political COVID-19 arguments across arenas

Researchers have developed a new corpus of 17,000 sentences focused on argumentation in German political communication related to COVID-19. This corpus spans three distinct arenas: plenary speeches, committee meetings, and press conferences. Contrary to initial hypotheses, the analysis revealed that arguments supported by domain-specific expertise were more prevalent in press conferences than in committee meetings. Additionally, a pilot study on automated identification of argumentative passages indicated challenges in precisely defining boundaries and noted that models are susceptible to confirmation bias. AI

IMPACT This research could lead to improved NLP models for analyzing political discourse and identifying biased argumentation.

RANK_REASON The cluster contains an academic paper detailing a new corpus and analysis of argumentation patterns. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.CL →

• arXiv
• COVID-19

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →