ClaimPT: A Dataset for Claim Detection and Fact-Checking

It contains 1,308 articles provided by the LUSA news agency, each annotated by two trained linguists following a dedicated scheme for claim detection. Annotations include claim spans, claimers, topics, stance, and temporal information.

The resource supports research in claim detection, fact-checking, and misinformation analysis in low-resource languages. **Access** The full dataset is available under a Data Use Agreement. To request access to this dataset please fill out the form (ClaimPT - Request Form) and send it to: **joao.a.castro@inesctec.pt** **License** creativecommons.org/licenses/by-nc-nd/4.0/ The project´s GitHub Repository provides further documentation, baselines and samples: github.com/LIAAD/ClaimPT **If you use the ClaimPT dataset please use the following citation:**Campos, R., Sequeira, R., Nerea, S., Cantante, I., Folque, D., Cunha, L. F., Canavilhas, J., Branco, A., Jorge, A., Nunes, S., Guimarães, N., & Silvano, P. (2025).