Rochester, N.Y. - Imagine being able to predict whether or not you will get the flu based on what and where people tweet.
An algorithm created by University of Rochester post-doctorate fellow Adam Sadielk does just that.
It tracks GPS-tagged tweets about feeling sick, then sees how many of that user's followers also tweet about being sick.
It factors in the location of the tweets and can predict 8 weeks in advance when people will get sick with the flu.
"We can detract that somebody met sick individuals in the library yesterday," Sadielk said. "So we can predict what kind of diseases they are going to catch from those people."
Sadielk says he wants to start a company based on this technology.
Insurance and pharmaceutical companies are already interested.