Photo: krakrakra515 via YouTube
Nothing lasts forever.Since 2008, Google Flu Trends was the reigning, amazing tool that allowed you to predict when your state (and more recently your city) was showing large volumes of flu-infected people.
They’re about to be dethroned by researchers at the University of Rochester.
A team led by postdoc fellow Adam Sadilek have figured out a way to use Twitter to predict when any given individual will come down with a cold even eight days before they show symptoms, New Science reports.
The group analysed 4.4 million geolocated Tweets in New York City for one month in 2010 that indicated a person was sick. They devised an algorithm that made sure the Tweets were about being physically sick (as opposed to being sick of the New York Knicks, traffic, etc.).
By tracking individuals who visited the same 100 by 100 meter cell within a given time window, they could accurately predict with 90 per cent accuracy if they will be sick in eight days, Sadilek wrote in his paper on the project.
A version of their application for New York City is already live. And here’s a visualisation of the app at work:
NOW WATCH: Briefing videos
Business Insider Emails & Alerts
Site highlights each day to your inbox.