Domains in Motion: A NLP Framework for Mapping Wind Energy Terminology

Wind energy constitutes a dynamic and interdisciplinary domain, shaped by technological innovation, evolving societal debates, and the interplay of diverse knowledge fields. Understanding its terminol- ogy is challenging, as concepts change quickly and overlap across different areas. This study introduces LUMEN, a computational method that uses Large Language Models (LLMs) and semantic embeddings to identify and organize subdomains within wind energy. Using a corpus of texts from 2016 to 2024 col- lected via Sketch Engine, LUMEN maps relationships among terms and highlights connections across the field. By providing translators and terminologists with a clear framework to visualize and navigate these complex terminological networks, the method demonstrates how NLP-based approaches can make the analysis of evolving, interdisciplinary domains both more precise and accessible.