Google Flu Trends mines aggregated search queries to estimate flu activity across much of the world.
7 Big Data Solutions Try To Reshape Healthcare
(click image for larger view and for slideshow)
This winter is turning into a severe flu season across large sections of the United States. According to the U.S. Centers for Disease Control (CDC), 29 states and New York City are reporting high levels of influenza-like illness (ILI), while another nine states are experiencing moderate flu levels.
What's the best way to monitor flu activity across the world? That's debatable, of course, but Google has an innovative solution: Use aggregated search data to track the flu in "near real-time," according to the company.
The Google Flu Trends site isn't new -- Google.org, the company's philanthropic arm, launched it in 2008 -- but it's a good example of how organizations and governments can mine big data for valuable insights.
So why use search queries to track flu activity worldwide? After all, isn't that what global health agencies like the CDC do already? Yes, but Google Flu Trends, by analyzing aggregated queries, can detect disease outbreaks much faster than these agencies, Google claims. And while health reports are often updated weekly and limited to a single country, Google Flu Trends has a near-global reach: It gathers data from wherever people use Google search. And since it's updated daily, it delivers more timely information.
"We have found a close relationship between how many people search for flu-related topics and how many people actually have flu symptoms. Of course, not every person who searches for 'flu' is actually sick, but a pattern emerges when all the flu-related search queries are added together."
By comparing query totals with data from conventional flu surveillance systems, Google has found that flu-related search queries are (not surprisingly) quite common during the flu season. And by counting the number of these queries, Google can then estimate flu activity in regions of the world that use its search engine.
Google determines flu activity levels -- intense, high, moderate, low or minimal -- by comparing current estimates from search data with official historic influenza information for a particular region. On January 8, 2013, for instance, it listed flu activity in the U.S. as "intense," a determination in line with the CDC reports of severe flu outbreaks across much of the country.
Flu Trends uses IP address information from Google's server logs to determine the origin of search users' queries.
Google doesn't position Flu Trends as a replacement for traditional data from health agencies, but rather as a complement that can help public health officials detect disease outbreaks early on, and hopefully limit the number of people affected.
In January 2008, for instance, Google Flu Trends detected a significant increase in flu activity in the U.S. Mid-Atlantic region. By comparison, published CDC reports were about two weeks behind, and hadn't yet shown this increase.
Conventional flu-surveillance reports typically come from doctors and health professionals. They're a good source of demographic data, which health authorities can't get from search queries.
Flu Trends' reach isn't truly global at this time; Google provides flu estimates for more than 25 countries from North and South America, Europe, Australia and parts of Asia. However, it doesn't include flu data for China, India, Indonesia, the Middle East and almost all of Africa (except for South Africa).
Of course, most search users don't want Google to keep track of every time they're (potentially) ill. The search giant addresses these privacy concerns by using aggregated, anonymized counts of weekly queries.
One thing Google Flu Trends can't do: Suggest the best recipe for chicken soup.
Predictive analysis is getting faster, more accurate and more accessible. Combined with big data, it's driving a new age of experiments. Also in the new, all-digital Advanced Analytics issue of InformationWeek: Are project management offices a waste of money? (Free registration required.)
6 Tools to Protect Big DataMost IT teams have their conventional databases covered in terms of security and business continuity. But as we enter the era of big data, Hadoop, and NoSQL, protection schemes need to evolve. In fact, big data could drive the next big security strategy shift.
Big Data Brings Big Security ProblemsWhy should big data be more difficult to secure? In a word, variety. But the business won’t wait to use it to predict customer behavior, find correlations across disparate data sources, predict fraud or financial risk, and more.