The chapter demonstrates how large vision-language models can support quantitative image analysis. It compares how parties in the German Bundestag used images to inform the public and evoke emotion during COVID-19 debates on Twitter.

Citation: Kessling, P., & Wiedemann, G. (2026). Computergestützte Bildanalysen mit Large Vision Language Models am Beispiel der Corona-Debatte auf Twitter. In F. Brand, H.-J. Bucher, & A. Schulze (Eds.), Handbuch Krisenkommunikation in der Pandemie (pp. 1–25). Springer VS.

Publication: https://doi.org/10.1007/978-3-658-47239-9_31-1