Tracking Flu in Real Time

Scientists and researchers at Johns Hopkins University and the Johns Hopkins School of Medicine scientists and researchers have developed a method of tracking actual U.S. flu cases by sifting thousands of publicly available Twitter posts per minute, according to a Jan. 24 report on the school's website by Phil Sneiderman. The team's screening method produces results in real time that track closely with CDC disease data taking much longer to compile, he writes.

This builds on a 2011 study by Mark Dredze, an assistant research professor in JHU's Department of Computer Science, showing that analyzing tweets can derive useful public health information. Dredze and his colleagues tested their system in November and December 2012 and showed that it matched up with CDC data much better than previous methods using tweets. They have created U.S. maps illustrating the much higher incidence of flu virus this winter than was occurring during the same period one year ago, according to Sneiderman's report, which is available at http://hub.jhu.edu/2013/01/24/using-twitter-to-track-flu.

Explaining that their method "looked only at public tweets in which all user names and gender information had been removed," he reports the team members hope to share their system with leading government health agencies. (Their research was funded partly by the National Institutes of Health's Models of Infectious Disease Agent Study.)

Posted by Jerry Laws on Jan 28, 2013


comments powered by Disqus