Chronic illness can be overwhelming to patients because it impacts many areas of their daily lives. Accordingly, many patients turn to online health support groups to get social support. In face-to-face patient support groups, health professional moderators provide clinical expertise within the context of peer-patients’ sharing of experience. However, in online health community settings, because health professionals’ time and resources are expensive, it is challenging to get health professionals’ opinions for thousands of messages posted each day. To solve this problem, I proposed to develop methods and techniques that maximize the use of already available clinical expertise online for online peer-patient conversation threads by developing a system, InfoMediator. The InfoMediator will semi-automatically weave health professionals’ existing answers to patients’ questions into peer-patient conversations by using Natural Language Processing (NLP) techniques complemented by user feedback.
Focusing on persons with diabetes, the outcomes of the proposed research will help us understand how to empower persons with diabetes to improve self- efficacy and self-care, while increasing the quality of online health information environment.
Click here for related publications to this project
PI: Jina Huh-Yoo
Mentor Chair: Wanda Pratt
Consultant(s): Amber Vermeesch (Nurse Practitioner), John Crowley (Alliance Health Networks)
NLP: Chun-Nan Tsu
Statistics: Jihoon Kim
Research assistant: Jing Zhang
Joyce Chai (NLP), Barb Given (Clinical research)
Student: Couri VanDam (NLP, machine learning), Shaheen Kanthawala (HCI, health informatics)
Please click here for project updates