Solidarity with Migrants/Women in German Political Debates: An Analysis via Large Language Models

Researcher

Aida Kostikova

Publications

• Aida Kostikova, Benjamin Paassen, Dominik Beese, Ole Pütz, Gregor Wiedemann, Steffen Eger (2024): Fine-Grained Detection of Solidarity for Women and Migrants in 155 Years of German Parliamentary Debates Full Publication

Collaboration

Sanne Hoeken, Sophie Spliethoff, Judith Sieker

Research Theme

R2 Prosilience & Robustness R3 Sustainability & Efficiency

Tags

LLMs

Our study explores the potential of Large Language Models (LLMs) in exploring the trends in solidarity with two groups — women and migrants — within German political debates in the last 155 years. By (1) manually annotating text for expressions of solidarity and anti-solidarity, we (2) train a classifier on our annotated data, with which (3) we infer large-scale trends.This exploration is twofold: we assess the capabilities of LLMs in social science tasks by evaluating their performance in tracking long-term trends of solidarity and anti-solidarity, and we aim to analyse these historical trends themselves.