Google Updates Flu Trends to Improve Accuracy

Updated on

Google Inc. is updating the software model it uses to estimate the spread of the flu to improve its accuracy, so that public health officials can quickly respond to outbreaks.

The company introduced Google Flu Trends in 2008 to help track the spread of the virus by using Web-search data. Now, Google will incorporate data from the Centers for Disease Control and Prevention in the U.S. to boost its accuracy, Christian Stefansen, senior software engineer at Google, said on the company’s research blog yesterday.

This is the second update to the flu model after the data for the U.S. “significantly overpredicted” flu levels in the 2012-2013 season amid heavy media coverage of the illness. The service showed better results last year, Google said, and the company is building on that ahead of this year’s flu season.

“It’s important that public health officials and health professionals learn about outbreaks as quickly as possible,” Stefansen said. “These models nicely complement other survey systems -- they’re more fine-grained geographically, and they’re typically more immediate.”

Google is trying to leverage its technological prowess and financial heft to address health problems and other societal challenges. The company’s philanthropic arm, which handles the flu service, also supports technology efforts to fight human trafficking, bolster wildlife preservation and encourage more education for girls.

The new flu trends model is for the U.S. and may roll out to other areas around the globe, depending on its success, Google said. Google Flu Trends now covers 29 countries, and the company also tracks trends on dengue fever in 10 countries.

Before it's here, it's on the Bloomberg Terminal. LEARN MORE