Google is now putting its massive power to work tracking disease outbreaks. Of course, it’s not really Google doing the work so much as it is the billions of people who type search queries into Google every day.
Researchers at Harvard University have successfully created a mathematical model to track the spread of infectious diseases using Google searches, as detailed in an article published in the journal “PLOS Computational Biology.” The Harvard study tracked dengue fever in Brazil, Mexico, Singapore, Taiwan, and Thailand.
Dengue fever is a mosquito-borne disease which afflicts around 400 million people per year. The symptoms of dengue fever include a high fever, nausea, headache, and skin rash. In the vast majority of cases, dengue fever isn’t fatal. It clears up after about a week, and there are generally no lasting aftereffects.
Dengue fever remains a persistent problem in many underdeveloped countries, where it can be difficult to track and combat outbreaks.
In the Harvard study, researchers tracked searches related to dengue fever on Google “Trends.” They also examined data about past outbreaks from health agencies. The idea is that when an outbreak occurs, there will also occur a corresponding spike in Google searches related to the symptoms of dengue fever.
These methods produced more accurate results than previous studies attempting to track disease outbreaks using Google data. The model worked best in countries with long histories of recurrent dengue fever outbreaks, and in countries where outbreaks occurred in high numbers.
The model produced the worst results in Taiwan, which has a short history of only a few years with dengue fever outbreaks.
Researchers are also using websites other than Google to track the spread of infectious diseases. Previously, researchers from Northwestern University modeled the spread of the flu using data from Twitter. They were able to predict flu outbreaks 6 weeks earlier than with other methods.