Large Language Models and Biomedical Annotators to Extract Information from Scientific Literature in Inflammatory Bowel Disease

Abstract

This poster discusses two automated approaches for identifying important information from scientific literature to assist literature reviews. A Large Language Model (LLM) and biomedical text annotator (UMLS-based MetaMap) were used to identify and ascertain keywords and methods from a collection of Inflammatory Bowel Disease (IBD) abstracts. The results were assessed by human experts for accuracy and relevancy.