During the V International Scientific and Practical Conference ‘Corpus Linguistics in Science and Education’, organised by the National Pedagogical Dragomanov University in Kyiv, the scientific community’s attention was focused on the plenary session delivered by Prof. Roman Roszko from the Institute of Slavic Studies of the Polish Academy of Sciences (ISS PAS). The presentation served not only as the substantive foundation of the event but also as a clear signal that Polish computational linguistics is entering a new digital era.
The presentation, titled ‘Evolution of the Team of Semantics and Computational Linguistics of ISS PAS: From Corpus Resources to Polish Language Models’, was met with immense interest from over 150 participants attending the session.
Prof. Roszko provided an exceptionally clear and profound outline of the journey taken by the current Semantics and Corpus Linguistics Team. This history evolved from building fundamental monolingual, bilingual, and multilingual corpora towards advanced solutions in the field of Artificial Intelligence (AI).
Key Pillars of the Presentation
The core of the lecture focused on the Institute’s involvement in projects critical to Polish technological sovereignty, including PLLuM (Polish Large Language Model), HIVE AI, CLARIN-PL-BIZ-Bis, and DARIAH-HUB.
Prof. Roszko highlighted the unique role of ISS PAS in creating high-quality datasets, specifically:
- Textual data for training, evaluation, and testing.
- Organic instruction data, which enables models to better understand user intent.
- Specialised programming data, essential for developing tools that support coding.
Another vital topic was the Institute’s role in verifying effective Text and Data Mining (TDM) reservations. In an era of mass AI model training, protecting resources and precisely managing data access has become a priority—a field in which ISS PAS serves as a leading expert.
Recognition from the Academic Community
The organisers from the Faculty of Foreign Philology at the Ukrainian State Dragomanov University expressed high praise for the contribution. In an official statement, they emphasised that Prof. Roszko’s speech was the ‘brightest highlight of the programme’, noting that the depth of the analytical and developmental work described, combined with the speaker’s high level of professionalism, set the tone for all subsequent scientific discussions.
‘This presentation caused a significant stir in the scientific community. We are impressed by the research depth and look forward to further fruitful cooperation within future scientific projects’, wrote the conference Organising Committee.
Prof. Roman Roszko’s appearance at the Kyiv conference confirmed that ISS PAS does more than just archive linguistic heritage; it actively shapes the future of computational linguistics, building bridges between traditional science and the technologies of tomorrow.
