Biomedical Annotation for Gut-Brain Interplay Information Extraction Meets Terminology Work: Insights and Perspectives

Within the context of the Conference and Labs of the Evaluation Forum (CLEF) 2025, the GutBrainIE @ CLEF 2025 challenge has been proposed. This natural language processing challenge represents one of the initiatives developed under the European-supported project HEREDITARY (HetERogeneous sEmantic Data integratIon for the guT-bRain interplaY). The focus of the challenge is on performing named entity recognition and relation extraction on a corpus of PubMed abstracts on the gut-brain interplay. The resulting annotations will be used to train information extraction systems, tailored to the biomedical domain under investigation. In this paper, we illustrate the concept systems that supported the development of the annotation schema, which served as the reference framework for both named entity recognition and relation extraction. Then, we examine the potential of reusing entity mentions and entity relations identified during the gold-standard dataset annotation process as terminological data in a medical terminology resource. Crucially, the task of named entity recognition differs from the process of term extraction that is adopted in terminology science to extract terminological data from textual documents. By examining annotated entity mentions and entity relations, we highlight how data obtained through named entity recognition and relation extraction can be partially reused for terminology work.