SPS22-109UL

Computational analysis of behavioral patterns underlying opioid misuse using social media

By: Ezra Player       

Department: Computer Science

Faculty Advisor: Dr. Rahul Singh

Opioid misuse is a significant public health crisis of our times. Finding high quality large-scale data for epidemiological modeling is particularly challenging for substance misuse. To address this challenge, our research focuses on the use of social media as a novel data source. We show that the anonymity and self-driven nature of discussions on social media lead to high-levels of disclosure and context-specific information. Further, based on computational extraction, mining, and analysis of content from the social media forum Reddit, we demonstrate how epidemiologically critical insights can be obtained.