This map indicates the prevalence of flu in the United States in the first week of January this year, 2013, as measured by the volume of flu-related tweets. States with higher flu rates are colored in a darker red. This year, in contrast to last January, the nation is awash in flu cases.
Credit: Mark Dredze/JHU
As if the coworkers’ sniffles, friends’ Facebook posts and frantic news headlines about this year’s flu season haven’t been enough to either educate or cause paranoia, the microblogging site, Twitter is also responsible for the heavy news feed about the illness no one wants to catch.
But, how do readers distinguish between the actual “news you can use” and people merely talking about the illness? Researchers at John Hopkins University’s School of Medicine have an answer.
Mark Dredze, an assistant research professor in the Department of Computer Science, along with his colleagues, created a new tweet-screening method to provide updates on the flu infection, but filter out unrelated conversations that are not useful to someone trying to be educated and aware.
The method filters about 5,000 public tweets per minute and, according to Dredze, the real-time data generated is closely paralleled with government disease data that may take a longer time to collect.
“When you look at Twitter posts, you can see people talking about being afraid of catching the flu or asking friends if they should get a flu shot or mentioning a public figure who seems to be ill,” Dredze said in a press release. “But posts like this don’t measure how many people have actually contracted the flu. We wanted to separate hype about the flu from messages from people who truly become ill.”
This strategy of using social media to track disease outbreaks has been gaining popularity in the public health sector. Filtering out tweets that indicate ill people can help estimate the severity of an outbreak. Last summer, the U.S. Department of Health and Human Services sponsored a contest for researchers who could come up with the best method to do this.
But, researchers must be wary of tweets that point to news reports or celebrities with illnesses, for they may provide a false outlook for the illness, said David Broniatowski, a School of Medicine postdoctoral fellow in the Department of Emergency Medicine’s Center for Advanced Modeling in the Social, Behavioral, and Health Sciences.
“For example,” he said, “a recent spike in Twitter flu activity was caused by discussions about basketball legend Kobe Bryant’s flu-like symptoms during a recent game. Mr. Bryant’s health notwithstanding, such tweets do very little to help public health officials prepare our nation for the next big outbreak.”
But, the John Hopkins researchers say that their method was created to combat this problem by using “statistical methods based on human language processing technologies,” stated the press release. This means their system can filter out messages about people who actually have the flu versus those who are worried about it.
“In late December,” Dredze said in a press release, “the news media picked up on the flu epidemic, causing a somewhat spurious rise in the rate produced by our Twitter system. But our new algorithm handles this effect much better than other systems, ignoring the spurious spike in tweets.”
Dredze hopes that the new system will be used by government agencies in the future.
“This new work demonstrates that Twitter posts can be used to guide public health officials in their response to outbreaks of infectious diseases,” Dredze said. “Our hope is that the new technology can be used track other diseases as well.”
____________
Other researchers in the study were doctoral student Michael Paul and recent bachelor’s degree graduate Alex Lamb, both in the Department of Computer Science.
View a video produced by Twitter about Johns Hopkins’ use of tweets to track public health trends here: http://www.youtube.com/watch?v=HmDIh-YS0GI
Layman's Terms Media 
Leave a Reply