Extraction and abstraction of lexicographic data through AI. The case of Body Part Nouns

This paper aims at exploring the use of Artificial Intelligence for lexicographic purposes. Focusing on a specific domain that is underrepresented in lexicographic works - namely, the domain of Body Part Nouns - the contribution employs AI to i) extract phraseological expressions and authentic examples in context, and to ii) organize the data according to specific properties, such as semantic domains and cognitive processes involved. The paper describes both the prompt-construction phase and the qualitative evaluation of the results obtained. Extraction tasks are compared with analysis tasks in order to identify the areas in which the LLM model achieves better performance and those where it still has limitations. While the primary focus of this contribution is on Italian data, the study also offers some contrastive insights into English.