This time of year, Twitter and Google searches capture innumerable complaints and questions about the flu. Luckily, new research established by Google and Johns Hopkins University claim this online buzz is a good thing. By using aggregate flu-related terms, you can now analyze social media trends to track and predict flu outbreaks.
You Are What You Tweet
Johns Hopkins computer scientists and researchers in the School of Medicine have crafted a tweet-screening program, called Ailment Topic Analysis Model (ATAM), to distinguish between individuals with the flu from those who are merely talking about it. With this data, the program can determine geographical areas where the influenza has become an epidemic. Additionally, the program can predict where the flu will become more prevalent.
Similarly, Google recently developed a program that predicts flu outbreaks based on popular Google searches. Google Flu Trends is able to track conversations in over 25 countries and provide real-time estimates of the number of flu cases in a specific geographic area. This means that every time an individual uses Google to search “influenza” or “fever,” his or her search will be incorporated into the Google Flu Trends dataset.
Dorm Room Diaries
This information could be especially pertinent on college campuses. As a college student, I am never more than five feet away from my smart phone. It’s no secret that my generation is particularly vocal on social media sites. We are the first to complain about a sore throat, but the last to seek in-person medical treatment from a primary care physician. We choose to be at the mercy of online symptom-checking applications.
College campuses have yet to get on board in a formal way, but The Center for Disease Control and Prevention (CDC) is one of the largest health agencies to incorporate this data into their relief efforts. Traditionally, local and state health agencies require weeks to prepare reports of flu activity; now flu data is available at your fingertips, with a few clicks of the mouse.
Public Profile
During a comparison between CDC flu estimates and the results of the ATAM screening program, Johns Hopkins found that ATAM yields the most accurate data. Although the Johns Hopkins team only had access to public Twitter profiles, potentially skewing data, researchers don’t believe this significantly influenced results.
How do you find information about the flu and other ailments? Do you head straight for your physician’s office, or do you search the web first?