Google Flu Trends is not sufficient as an early epidemic detection system but could be if augmented with additional computational intelligence, according to researchers publishing this week in the Journal of Medical Internet Research.
The researchers, based at Lahore University of Management Sciences in Pakistan, acknowledge that Google Flu Trends serves as a good "baseline indicator" of epidemic trends but dismissed the notion that it could, for example, serve as a good warning system for emergency departments. To enhance the system, however, they built upon the system, developing a new system dubbed FluBreaks, which converts Google Flu Trends data into an early epidemic detection system through the use of "sophisticated statistical analysis."
"Our research is the first attempt of its kind to relate epidemic prediction, using Google Flu Trends data, to Internet penetration and the size of the population being assessed," the study's authors wrote. "We believe that understanding how these factors affect algorithms to predict epidemics is an integral question for scaling a search query-based system to a broad range of geographical regions and communities."
According to the researchers, their system helps Google Flu Trends to move past traditional approaches used by the Centers for Disease Control and Prevention for identifying epidemic trends such as cumulative sum distribution and historical limits method distribution.
"While we did not find a single best method to apply to Google Flu Trends data, the results of our study strongly support negative … algorithms being more useful when dealing with potentially noisy search query data from regions with varying Internet populations," the researchers wrote.