Population Research Discovery Seminars
Annotating Social Determinants of Health Using Active Learning, and Characterizing Determinants Using Neural Event Extraction
Kevin Lybarger, Postdoctoral Fellow, UW Biomedical Informatics and Medical Education
Register for Zoom Seminar HERE
Social determinants of health (SDOH) affect health outcomes, and knowledge of SDOH can inform clinical decision-making. Automatically extracting SDOH information from clinical text requires data-driven information extraction models trained on annotated corpora that are heterogeneous and frequently include critical SDOH. This talk will present a new corpus with SDOH annotations, a novel active learning framework, and the first extraction results on the new corpus. It will also describe an up-coming NLP shared-task that will utilize this corpus.
Kevin Lybarger is a Postdoctoral Fellow at the University of Washington in the Department of Biomedical Informatics and Medical Education. His research interests combine data-driven machine learning and natural language processing (NLP) with important real-world problems. His current research explores the intersection of NLP and clinical informatics, including the extraction of information from clinical text that can improve healthcare and advance clinical research. Dr. Lybarger holds a PhD in Electrical & Computer Engineering from the University of Washington.