The Google Press office today brought to my attention an unusual and totally unexpected bit of Google trivia: the modelling of real-world phenomena using patterns in search to map the spread of influenza across the US.
Google is proving to be far more than just a search engine these days. In fact, today it repositioned itself as a time travel machine which allows users of Google Earth to travel back to 320 AD and explore more than 6700 buildings as they stood at the centre of the Roman Empire back then. The biggest collection of 3D buildings on the Google Earth map, together with the input of noted historians, makes this possible.
Anyway, back to the flu thing. It is really an extension of the Google fascination with trends in online search queries. we saw it with the recent US Presidential election, we have seen it with the annual Zeitgeist 'what's hot' listings and now we see it with patterns of influenza distribution.
In many ways, this latter trend mapping is more important, if a lot more boring, than the others. I doubt many casual users could really care less that more people in Maine have the flu than they do in New York, for example. Yet influenza is responsible for thousands of deaths every year, as many as half a million worldwide in actual fact. By comparing aggregated search queries during 'flu season' with data supplied by the U.S. Centers for Disease Control and Prevention (CDC) the Google engineers were able to determine that there is a close relationship between the frequency of the search queries and the number of people experiencing flu-like symptoms each week.
"As a result, if we tally each day's flu-related search queries, we can estimate how many people have a flu-like illness" a Google spokesperson comments. Now anyone can tap into the work that the team has been doing courtesy of Google Flu Trends, with up-to-date flu activity estimates for all 50 US States.
Why should you care? Well maybe because by making these estimates available on a daily basis, instead of the week or two it typically takes traditional flu surveillance systems to produce reports, it might just save some lives.