LINGUA PATRIA — Open Subset v1.0

LINGUA PATRIA is a trilingually aligned parallel corpus covering African constitutional governance, anti-corruption, and patrimonial sovereignty. This Open Subset v1.0 contains 77 segments (~21,000 words per language) in English, French, and Portuguese, extracted from "The Patrimonial Revolution: Public Goods & Africa's Third Independence" (Armand Salouo, Kurukan Publishing, 2026). This dataset is the open-data foundation for GOVBENCH AFRICA, a multilingual NLP evaluation benchmark assessing language model performance on African governance tasks in Twi/Akan, Wolof, and Lingala (v2.0, in development). Responsible Use Policy available on transparencyafrica.org/ressources.html Note: This is the DPG-nominated artefact (Digital Public Goods Alliance). The full 80,000-word LINGUA PATRIA corpus (CC BY-NC 4.0) is maintained separately by Kurukan Publishing. This open subset is independently licensed under CC BY 4.0 and constitutes the complete, self-contained Digital Public Good. Related resources:- Working papers (Zenodo): https://zenodo.org/search?q=creators.name%3A%22Salouo%2C+Armand%22- Author ORCID: https://orcid.org/0009-0000-5155-5856- Contact: editor@cmpublishing.online