Interpretable ML project combining Global Terrorism Database event records with international news framing to analyze terrorism severity rankings, residuals, and feature importance.
-
Updated
May 2, 2026 - Python
Interpretable ML project combining Global Terrorism Database event records with international news framing to analyze terrorism severity rankings, residuals, and feature importance.
Zero-shot news framing classifier (BART-MNLI) validated with an independent LLM judge across Reuters, Fox News, and Politico articles. Includes a documented case study of where zero-shot labeling breaks down.
Comparative analysis of how domestic Ethiopian and Western media framed the Tigray conflict (2020-2022), with data and reproducible methods.
Media framing of 4chan in Canadian government-funded media (2015-2025) — a reproducible BERTopic pipeline
To associate your repository with the media-framing topic, visit your repo's landing page and select "manage topics."