Strolr: An LLM-enabled Chatbot to Support Pregnant Women’s Quick and Easy Information Seeking from Trustworthy Sources

Presentation Time: 04:30 PM - 05:00 PM

Abstract Keywords: Delivering Health Information and Knowledge to the Public, Large Language Models (LLMs), Personal Health Informatics, User-centered Design Methods
Primary Track: Applications
Programmatic Theme: Consumer Health Informatics

Despite the vast knowledge available online, pregnant women face numerous challenges in health information seeking, such as low trustworthiness, poor understanding, and an overwhelming number of resources. To address this, we developed Strolr – a large language model-enabled chatbot that retrieves and summarizes trustworthy governmental resources. In the demonstration, we aim to highlight its ability to answer questions regarding common pregnancy-related health topics with links to cited sources in an easy-to-understand chat format.

Speaker(s):
Nidhi Soley, PhD Biomedical Engineering
Johns Hopkins University

Michelle Nguyen, BS
Johns Hopkins University School of Medicine

Ilia Rattsev, PhD Candidate
Johns Hopkins University

Author(s):
Casey Taylor, PhD - Johns Hopkins University; Angie Jelin, M.D. - Johns Hopkins School of Medicine;