(https://www.instagram.com/p/DdVrMV3GnC2/?utm_source=ig_web_copy_link&stkn=MzRlODBiNWFlZA==)
A new version of ParaRook (https://uacorpus.org/poshuk-u-graku/paralelni-korpusi-pararook) has been released — a family of parallel corpora based on GRAK (the General Regionally Annotated Corpus of the Ukrainian Language).
In the new version of ParaRook:
• the English, German and French-Ukrainian corpora have been significantly expanded (the EN→UK corpus contains over 15.3 million tokens, and the DE→UK corpus 13.5 million);
• new language pairs have been added: Arabic, Belarusian, Hungarian, Italian, Swedish and Turkish. In total, 15 languages and a collection of translations of the Qur’an are covered;
• the morphosyntactic annotation of the Ukrainian section has been updated (using the UDPipe2 model on the UD Ukrainian ParlaMint Treebank);
• A feature highlighting translation equivalents in the interface has been launched on an experimental basis to facilitate rapid analysis of translations.
Lecturers, researchers and students from the National Technical University ‘Kharkiv Polytechnic Institute’, Kyiv National University, Lviv National University, the Ukrainian Catholic University and the National University of Kyiv-Mohyla Academy joined the corpus compilation and annotation efforts between 2023 and 2026.
