February 8, 2013
Researchers Use Twitter Posts To Track Flu Rates
[Watch Video: Tracking The Flu With Twitter]
Enid Burns for redOrbit.com — Your Universe Online
If you tweet when you feel under the weather, your 140 characters may just become an indicator for a flu outbreak in your area. Researchers at the University of Rochester in upstate New York have used the GPS tags on Twitter to map out areas in cities where the flu rate is higher, or nonexistent.
Over the past year researchers looked at several metropolitan areas, primarily in the U.S., to identify patterns in chatter on Twitter.
"If you want to know, down to the individual level, how many people are sick in a population, you would have to survey the population, which is costly and time-consuming," said Adam Sadilek, postdoctoral researcher at Rochester. "Twitter and the technology we have developed allow us to do this passively, quickly and inexpensively; we can listen to what people are saying and mine this data to make predictions."
Many tweets are geo-tagged, and carry GPS information that shows precisely from where a tweet was sent. Algorithms can track that data, along with context, and compare it to other tweets of similar topics made from the same geographical area. That tweet, "Feeling sick and leaving work early," could help map out where in a particular city the flu is denser. The data can then be used to get more medicine to that area, or for medical staffs to put more workers on duty.
Individuals can also use the data. Sadilek said someone might use the data to determine whether to get onto a particular bus. It can certainly be useful for an individual to determine where to go pick up groceries or conduct other activities.
The system developed by the University of Rochester shows data in real time. The key is that it collects data using Twitter. Tweets from individuals are basically self-reporting, and public. It isn't digging into semi-private posts on Facebook or other social networking platforms. People are sharing information on Twitter and it is collected into the database to form maps and correlations.
This is not the first time Twitter has been implemented to track trends, including illness.
Twitter is often used to identify trending information, and an algorithm developed by MIT researchers helps dig into the data. On the health front, recent research released by Johns Hopkins University used the micro-blogging social network to track the spread of infectious diseases, including the flu. And the U.S. government is using Twitter to track health issues on a local level.
The Rochester team hopes to make this data useful to metropolitan areas, and individuals within those areas. They are developing an app that color-codes users according to their health by mining information from tweets in 10 cities worldwide.
"This app can be used by people to make personal decisions about their health. For example, they might want to avoid a subway station if it's full of sick people," Sadilek suggested in a University-released statement. "It could also be used in conjunction with other methods by governments or local authorities to try to understand outbursts of the flu."