Mining and Mapping the Swine Flu

The AP reported yesterday that Kirkland, Washington-based biosurveillance firm Veratect detected the swine flu outbreak using data mining. And told the Centers for Disease Control about it 18 days before the World Health Organization issued its first warning.

According to the story,

Veratect, based in Kirkland, Wash., uses a technique known as "data mining" to automatically search tens of thousands of Web sites daily for early signs of looming medical problems or civil unrest anywhere in the world. Anything of interest is turned over to a team of 35 analysts to determine its significance and to post on the company's Web site.

The story later brings in its expert:

"This approach is not yet vetted," said Dr. Marguerite Neill, an infectious disease specialist at Brown University and a spokeswoman for the Infectious Disease Society of America. "It is an interesting idea, but we haven't used it before."

Sounds an awful lot like, "but we've always done it this way" to me.


And in other data-related flu news, the mysterious niman has been outed as Henry Niman, a biochemist in Pittsburgh, in CNN's punnily-titled "Online swine flu map goes viral." Niman is the creator of the flu map I posted here on Monday. CNN linked to the Google Maps Mania blog, which promises a collection of swine flu map links, but of course, that blog is now overloaded with traffic.

So, I'll leave you with HealthMap.org, which appears even more outdated than Niman's. But which is accepting submissions, junior virogeologists.