Skip to main content

University of Michigan School of Information


Mining social media to characterize community health

This project will advance data-driven methods to generate localized and community-centric insights, using new social media sources, as a means of accelerating progress towards community health equity. Specifically, we will develop and validate social media-based measures for the two important health behaviors driving negative health outcomes and disparities in the US: physical activity and alcohol consumption using the spatially referenced activity of users of the social media websites. We hypothesize that social media-based measures can accurately characterize health behaviors, attitudes, and resources — all at the census tract level. Moreover, we hypothesize that these measures can be effectively derived from social media for areas with smaller populations. Data such as those used in our research represent a profound opportunity for community health disparity reduction efforts, since they are available at a larger scale and finer resolution than has previously been available, and much of it is regularly refreshed.

Principal Investigator: Vinod Vydiswaran

Co-investigators: Tiffany Veinot, Daniel Romero

Start Date:
End Date:

The amount of the award is $60,000 for UMSI for the project period. The grant is funded by the Endowment for the Basic Sciences.