NONWESTLIT builds two linked research databases and releases open datasets for computational research on historical, low-resource languages.
Explore the databases and open datasets behind NONWESTLIT’s comparative research.
The database contains articles from nineteenth-century Ottoman, Russian and Japanese periodicals, with metadata, transcriptions and expert tagging. As of September 2026, it holds over 23,000 documents.
The Translations Database records fiction translated into Ottoman Turkish, Russian and Japanese, linking translations to their originals, editions and paratexts. The database is in progress.
NONWESTLIT provides multi-level, multi-label text classification datasets in Ottoman Turkish and Russian. They are publicly available on Hugging Face.