AI Sport Science (LLM): A Terminological Framework for the Interface Between Large Language Models and Sport-Scientific Knowledge Systems

Priority registration for the concept of AI Sport Science (LLM) within the AUGMANITAI framework (Ehstand 2025–2026). Author ORCID: 0009-0006-3773-7796. AI Sport Science (LLM) designates the systematic study of how large language models represent, transform, and fail to represent sport-scientific knowledge — and the terminological infrastructure required to bridge the conceptual gap between sport science vocabulary and LLM interaction paradigms. Addresses the near-absence of LLM-specific research despite rapid AI adoption in sport, with SportQA (Xia et al. 2024, NAACL) showing 40% of LLM errors stemming from conceptual deficits. Full definitions maintained in the unpublished AUGMANITAI Compendium. ISO 704 / ISO 1087 / ISO 30042 compliant.