Predicting the Next Outbreak

The project will use a variety of social media data describing greater Boston—including Twitter posts, Craigslist listings, and Yelp reviews—in conjunction with infection rates to identify “signals” that might indicate the emergence of an outbreak of COVID-19. This will be the basis of an automated real-time warning system that could support public health officials to respond proactively. This will be especially as we move into a new phase of the pandemic in which the expansion of vaccination has diminished the virus’ prevalence but not yet established herd immunity, meaning outbreaks will be sporadic and less predictable.