Lviv University Linguists Unlock the Power of AI: Training on Large Language Models Completed in Jena, Germany

Four lecturers from the Faculty of Philology at Lviv University have successfully completed the training and met the requirements of the academic and pedagogical course “Large Language Models for Linguists,” organized by Friedrich Schiller University Jena (Germany). The course was part of the DAAD-funded programme “Ukraine Digital: Ensuring Academic Success in Times of Crisis (2024–2025)”.

Participants acquired advanced competencies and practical expertise in using large language models (LLMs) as indispensable tools for organizing and analyzing linguistic data. They mastered prompt and code writing for various theoretical and applied domains such as language teaching, translation studies, discourse analysis, grammar research, lexical semantics, and text digitalization. Each participant developed and defended a scientific project based on the use of large language models.

Dr. Olha Maksymiv, Head of the Department of Oriental Studies named after Prof. Ya. Dashkevych, presented a project on using LLMs to compile frequency dictionaries and compare their data. She conducted an experiment to complete the lemmatization of a Persian newspaper frequency dictionary without full access to the original concordances. Routledge’s Persian frequency dictionary data were converted into a comparable format, and Ukrainian corpus processing was initiated using an LLM with further refinement in a programming environment. The project concluded that while LLMs excel in performing technical tasks, they require philological oversight for nuanced linguistic work. A philologist must maintain control at every stage—large tasks cannot be delegated blindly to the model.

Dr. Solomiia Buk, Professor of the Department of General Linguistics focused on the lexical base of the Ukrainian language through the lens of large language models, easy-to-understand language, and its linguodidactic potential. In her presentation, she demonstrated the specific use of Scholar-GPT in linguistics, examined AI’s understanding of the normative documents that regulate the assessment of Ukrainian as a foreign language (particularly the CEFR), and analyzed the scientific principles for determining lexical and grammatical minimums for A1–C2 levels. Her results have significant potential for teaching Ukrainian as a foreign language, defining basic vocabulary for state standards, and broader applications in digital linguistics.

Oleksandra Antoniv, Senior Lecturer at the Department of Ukrainian Applied Linguistics shared insights from her final project on working with Ukrainian phraseological material in a foreign language audience. In addition to linguodidactic approaches to selecting stable linguistic units for foreign students, she proposed various types of tasks that teachers of Ukrainian as a foreign language can create using large language models such as ChatGPT, Gemini, and ElevenLabs. This material can be particularly valuable in classes for students of the “Ukrainian Language and Literature. Ukrainian as a Foreign Language” educational program. It aims to foster critical thinking, demonstrate the advantages and disadvantages of using artificial intelligence in task creation, highlight “hallucinations” of large language models, and promote the development and improvement of future educators’ digital literacy.

Oksana Shchupakivska, Senior Lecturer at the Department of Oriental Studies, explored the potential of LLMs in identifying proper names in Ukrainian and Arabic texts and interpreting them as precedent names – cultural codes, carriers of stereotypes, symbols, and associative-connotative meanings. The results of step-by-step testing of the model’s capabilities and corpus analysis of over 6,000 contexts from Sketch Engine confirmed the relevance of a hybrid approach – combining automation with expert interpretation. This opens up new prospects for innovative digital methodologies in intercultural linguistic research.

The online course was delivered in a blended synchronous-asynchronous format and included lectures, practical classes, and self-study, altogether comprising 4 ECTS credits.

The programme was made possible thanks to Dr. Nataliia Cheilytko, postdoctoral researcher at Jena University, with the support of Prof. Dr. von Waldenfels, head of the project “Network for Ukrainian Studies: Courses in Digital Ukrainian Philology 2024/2025” (Institute of Slavic and Caucasus Studies, Friedrich Schiller University Jena).
More details about the programme can be found at the official website: https://www.ukr.uni-jena.de/uk/30/navchalni-proyekti

This international training initiative not only enhanced the participants’ digital and linguistic competencies but also opened up new directions for interdisciplinary research. The acquired skills will significantly contribute to the modernization of Ukrainian linguistic education and the integration of AI technologies into academic practice.

Close cooperation between the Philological Faculty of Lviv University and Jena University would not have been possible without Dean Dr. Roman Krokhmalnyi’s support for the idea of international cooperation and mutual recognition of diplomas in related academic disciplines. Students from our faculty also participated in numerous online courses as part of the 2022–2025 projects.

   S.Buk

Certificate_LLM_Oksana Shchupakivska Certificate_LLM_Oleksandra Antoniv Certificate_LLM_Olha Maksymiv Certificate_LLM_Solomiya Buk (1)