Everyone has taken a wrong turn in a strange city and suddenly come upon what seems like a sketchy neighborhood. But is that gut feeling an accurate one?
MIT researchers have created a research tool using Google Street View to crowdsource people's snap judgments of an urban place, based on just one picture, as wealthy or poor, safe or dangerous, lively or boring. They hope the data can be used to study the relationship between public perception and actual crime rates, affluence, or other indicators of livability.
Their dataset contains over 100,000 images taken within the city limits of 56 cities around the world. Some surprising patterns have arisen from the first 2 million clicks—Washington, D.C., apparently appears safer than New York City, and Tokyo more depressing than Minneapolis—but the researchers hope that volume will smooth out the noise.
In the original iteration of the experiment, comparing Boston and New York City to the Austrian cities of Linz and Salzburg yielded evidence that high rates of American social inequality are easily visible on the street, and that they do indeed correlate with murder rates for the two cities.