With flu season upon us, Google is trying to keep you healthy. They’ve recently updated their Google Flu Trends formula, which makes weekly predictions on flu cases around the world using search indicators. When it first launched in 2008, this product seemed promising and a prime example of how big data can be used for social good. However, since launching, the data has been off. In the 2012-2013 season, Google overshot by 95%. And the following year by 75%.
It takes a big (gigantic) company to admit when they are wrong, and Google has been working since to figure out a more accurate algorithm. The new algorithm is just as mysterious as the old, but it will integrate CDC data continually with Google’s search trends on both the flu and dengue fever. This area of research is very new, and so is the idea that big data doesn’t have to be this scary thing — it can actually help protect people and even give them the info they need to stay healthy.
The Wall Street Journal uses Global Pulse, the United Nation’s big data project, as an example. Global Pulse crawls through Twitter data to identify potential international crises before they happen. They are also working with a large telephone company to find patterns surrounding the spread of diseases like malaria and HIV, as well as determining if mobile phone data can provide more accurate poverty indicators.
There’s still plenty of work to be done, and we’re anxious to see how Google’s improved Flu Trend effort fares this year.
How would you use big data for good? Are you flu-phobic? Sound off below!