Improving Laypeople Familiarity with Medical Terms by Informal Medical Entity Linking

Abstract

Social media is an important community engagement tool in the health domain for health promotion, patient education, and outreach. Considering its educational potential, we introduce an Informal Medical Entity Linking (EL) model with three components to enhance the laypeople comprehension of medical terminology by linking popularized medical phrases in social media posts to their specialized counterparts and to relevant Wikipedia articles. Medical experts assessed the accuracy and relevance of the EL model, finding that the Medical Concept Normalization (MCN) component of our model correctly classifies 89% of informal phrases. The second component, Entity Disambiguation (ED), effectively predicts relevant Wikipedia articles for identified popularized medical terms, and the third one, Learning to Rank (LTR), improves the identification of the most relevant Wikipedia articles based on popularized medical phrases and their specialized counterparts. These findings suggest our model can be a valuable tool for enhancing laypeople’s comprehension of medical terminology by leveraging social media’s educational potential.