Imagine if city inspectors could know in advance which restaurants are likely to have food safety violations, and then swoop in and correct those problems before customers’ health is jeopardized.
That’s the dream outcome of the latest project by DrivenData, a year-old Boston startup that helps nonprofits and government agencies use data science to address social problems. It’s running an online contest that invites anyone with the technical know-how to crunch two data sets -- City of Boston food violation inspection reports and Yelp restaurant reviews – and predict where future hygiene violations might occur.
Later, once new city inspection data are available, contestants will learn whose predictions came true. The best guesser will win $3,000, and two runners-up will receive $1,000 apiece.
“If food inspectors know where the concerns might be as reviews get posted in real time, then they can get to problems faster and go where they’re most needed,” said Harvard alum Isaac Slavitt, who cofounded DrivenData with fellow alum Peter Bull and Harvard Business School student Greg Lipstein. The company is supported by the Harvard Innovation Lab’s Venture Incubation Program and the Rock Accelerator Program at HBS.
“This shows cities what can happen when they open up the data they have,” Slavitt added, “and it shows data scientists what sorts of things they might be able to help the public sector with.”