Joho the BlogGoogle flu interview - Request for Help - Joho the Blog

Google flu interview – Request for Help

I’m going to be on the radio news show Here and Now tomorrow to talk about Google.org’s ability to track outbreaks of flu by charting search terms (“flu symptoms”), time, and presumed IP location. I plan on talking about it as an example of the power of having enormous amounts of data, and of putting to use information generated for some other purpose.

Any ideas about how else this sort of technique could be used or is being used? (Amazon’s personalization is a different sort of example.) Any concerns (other than the how-not-to-do-it example from AOL)? [Tags: ]

11 Responses to “Google flu interview – Request for Help”

  1. […] ( Amazon ’s personalization is a different sort of example.) Any concerns (other than the how-not-to-do-it example from AOL)? More […]

  2. When I was at Microsoft, we used search terms to guide the programming “rhythm” of the Microsoft.com home page, promoting more home user content on Fridays and weekends and IT pro/dev traffic primarily during the work week.

    We also used it in real time during outbreaks of worms like Blaster and SQL Slammer, which may be closer to your epidemiological example. What we learned during that outbreak is that you need to track people’s queries on symptoms rather than diseases. Queries related to the names of Internet worms tended to track the news cycle pretty closely, while queries about symptoms of the infection preceded the news cycle and lingered long after everyone else was done talking about the problems.

  3. I’m not sure if this applies, but we did an analogous thing when redesigning the website at the McLuhan Program back in 2001. We designed the front page, and every link was broken. Then we waited for people to tell us that a link was broken. That interest indicated where we should address the development effort. Some pages that were never queried, we changed for more interesting content.

    The basic question is, what is capturing people’s attention, and how is that knowledge useful to understanding the reason said attention is being captured? Both human and systems epidemiology is one application. Market research (being done for a while in this way) is another. Social program spending in government policy could be another – for example, collecting data on searches related to teen pregnancy, birth control, abortion, morning-after pills, etc., could indicate geographical areas to which investments in reproductive education might be directed (and possible under an Obama administration. Yay!). Similar studies relative to mortgage foreclosures, credit counselling, etc., might help focus public policy attention more effectively.

  4. Anonymously collected location information in the form of which cell phone towers you were connected to would be very useful in things like dynamic traffic analysis and real time directions (“don’t take I 64 because the traffic is so heavy the risk of an accident and delay is also greater”).

    On the shadier side of things I imagine that there are multiple opportunities for near insider trading information leaks via search. If you had access to internal google search trends and saw a massive increase in “Buy Stock in XYZ company”, you are in a position of great opportunity.

  5. Here’s a great visualization of the TwitterVoteReports:

  6. Trying a comment box-acceptable version:
    votereport.us/TimeView/applet/index

  7. For a general treatment on this subject in a layperson-accessible book, see “Supercrunchers: How Anything Can Be Predicted,” by Yale Law School’s economist Ian Ayres, http://islandia.law.yale.edu/ayres/

  8. Take a look here: http://volokh.com/posts/1226438263.shtml

  9. – ten x ten
    is a 100 picture grid, each associated to a news article and to 1 tag
    it s the 100 most used tag on all major press agencies
    updated hourly and with a hell of an archive

    http://tenbyten.org/10×10.html

    – mark, ciao

  10. Probably too late to be useful, but I find this marvel quite perplexing.
    1) How do we feel about the fact that Google would know more than the CDC about the spread of an epidemic, and though it philanthropically would share that information with all due privacy safeguards, is under absolutely no obligation to do so, or not capitalize on its value by delaying a few hours to make some shrewd investments?
    2) If the nascent epidemic were not a coarse-grid phenomenon like influenza, but (let’s say) smallpox and it was inexorably spreading from house to house centered at David Weinberger’s residence, wouldn’t we want Google NOT to honor its principles of privacy and confidentiality when it noticed that people in a small area of Brookline were searching for information about rashes and blisters?

  11. cool pics


Web Joho only

Comments (RSS).  RSS icon