Harnessing AI for Creating Definitions of Georgian Military Terminological Neologisms in the context of the Russia-Ukraine War

The outbreak of the Russia-Ukraine war was followed by a rapid increase in drone-related terminology in the Georgian language. However, these terms remain absent from existing bilingual and encyclopedic Georgian military dictionaries to date. This paper explores whether Generative AI tools can be integrated into the terminological workflow, particularly to streamline the process of crafting draft definitions for Georgian military neoterms. For this, 10 drone-related multiword terminological neologisms were selected from Georgian comparable corpora, MiliCorpKa, consisting of analytical blog materials regarding the Russia-Ukraine war. Next, draft definitions were generated with five LLMs (ChatGPT, Grok, Claude, Gemini, DeepSeek) using two prompting strategies: a so-called basic prompt and an ISO-based prompt. The outputs were assessed against a self-developed evaluation scheme. Preliminary results of this study emphasize the importance of fine-tuning prompts and the differences in effectiveness among 5 models for creating definitions in Georgian. The results of this exploratory, qualitative research offer a glimpse of the extent to which Generative AI Chatbots can be a helpful tool for post-editing terminological work and for creating draft definitions for low-resourced languages, such as Georgian.