Project
GenAI-GDELT
Data pipeline, dataset, and companion site for the generative-AI governance framing study
End-to-end data-science pipeline that mines GDELT 2.0 via BigQuery, scores ~1.1M multilingual articles across six governance frames with a LaBSE-validated matcher, and ships the dataset plus an interactive companion site.
The engineering side of the Framing Generative AI Governance in Online News study: a reproducible data-science pipeline that extracts generative-AI governance coverage from GDELT 2.0 via BigQuery, applies a two-stage multilingual matcher (keyword pass + LaBSE embedding validation), and scores each article across six governance frames.
The project ships three artifacts — the code, a Hugging Face dataset of ~1.1M scored articles, and a deployed companion site that visualizes the frame trends and the policy-milestone event study.
End-to-end data-science pipeline that mines GDELT 2.0 via BigQuery, scores ~1.1M multilingual articles across six governance frames with a LaBSE-validated matcher, and ships the dataset plus an interactive companion site.
The engineering side of the Framing Generative AI Governance in Online News study: a reproducible data-science pipeline that extracts generative-AI governance coverage from GDELT 2.0 via BigQuery, applies a two-stage multilingual matcher (keyword pass + LaBSE embedding validation), and scores each article across six governance frames.
The project ships three artifacts — the code, a Hugging Face dataset of ~1.1M scored articles, and a deployed companion site that visualizes the frame trends and the policy-milestone event study.
Related Publication
- publication Framing Generative AI Governance in Online News