Online Program

Return to main conference page
Tuesday, January 7
Tue, Jan 7, 9:00 AM - 10:45 AM
Porthole
Patient-Centered Outcomes

Understanding patients’ feedback through natural language processing and machine learning (306771)

Presentation

*Yuhao Liu, Center for Helath Workforce Studies 

Keywords: patients' feedback, natual languange processing, machine learning

Understanding patients’ feedback is crucially important for hospital administrators. A large and increasing amount of unsolicited feedback on crowd-sourced review websites today makes information required for such attempts more accessible. The purpose of this study is to better understand patients’ concerns and expectations by analyzing their feedback collected from the internet.

This study retrieved 26,141 ratings and comments for 5,888 hospitals from Google Maps. Comments were classified into positive, neutral and negative in terms of their ratings. The top 20 most frequent two-gram terms were identified for positive and negative comments. Text clustering analysis for negative comments and sentiment analysis based on all comments were performed using four different machine learning algorithms.

This study suggests that hospitals may wish to reduce waiting time, improve billing and payment processes, make facilities accessible for aged visitors and improve service quality. Patients’ feedback on the internet is a useful and informative data source. Natural language processing and machine learning techniques help us classify and understand patients' feedback more efficiently.