Using NLP to Infer Patient Preferences from Social Media Conversations
Supported by a pilot grant from the BC SUPPORT Unit’s Health Economics and Simulation Modelling Methods Cluster, our team of health researchers and computer scientists is developing a natural language processing system that can help users to explore the preferences, needs, and experiences expressed by patients on social media like Reddit.
Our vision is that this software will make it easier for health system stakeholders to explore and take into account patients’ preferences at all stages of the health technology life-cycle.
For example, many pharmaceutical and biotechnology companies are using discussions on social media to help identify unmet need and inform drug development decisions. However, coding these discussions by hand takes a lot of work, and can constrain companies’ ability to explore patients’ perspectives on an ongoing basis.
Our system solves this problem by automatically analyzing social media discussions and making it easy for users to explore key themes and trends in patients’ preferences using an interactive dashboard.