Recent publication highlight : "A Neuro-Symbolic Method for Understanding Free-Text Medical Evidence" (article link) In this work, we proposed a new Self-Attention -- Medical evidence Dependency (MD)–informed Attention, by combining neural and symoblic approaches, for understanding free-text clinical trial publications with better generalizability and interpretability. If you are interested, here is an 3-min YouTube video I made to present the main ideas and results of it.
Me: I'm a Machine Learning Scientist at Tempus Labs, Inc., and base in New York City. I got my Ph.D. in Biomedical Informatics at Columbia University in 2021 (DBMI alumna! 🎓), with a focus on biomedical Natural Language Processing. (advisor: Dr. Chunhua Weng).
My doctoral dissertation -- "Towards a Unified Evidence Computing from Clinical Research Literature for Evidence-based Medicine"(link), was focused on developing novel language technologies to help doctors comprehend clinical research literature and facilitate the practice of Evidence-based Medicine at the point of care. (Click here for my CV)