LLMs analyze 18M tweets to map negative campaigning across Europe

Researchers have developed a method using large language models (LLMs) to classify negative campaigning across different languages, analyzing 18 million tweets from parliamentarians in 19 European countries. The study found that governing and coalition-oriented parties face greater reputational constraints against negative campaigning, while opposition and outsider parties have fewer. The research also indicated that parties further from the ideological center, particularly on the radical right, tend to employ more confrontational rhetoric. AI

IMPACT Demonstrates LLMs' capability for large-scale cross-lingual analysis, potentially transforming political science research.

RANK_REASON Academic paper detailing a new methodology and its application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• alphaXiv
• arXiv
• CatalyzeX
• DagsHub
• Gotit.pub
• Hugging Face
• large language models
• ScienceCast
• Victor Hartman

• paper
• policy

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