Tag
This paper presents an LLM-based pipeline for analyzing media bias and framing in online news, tested on Albanian articles with moderate agreement in annotations compared to automated methods.