Forgot your password?
Social Networks Stats Twitter Science

There Is a Fly In My Tweet: Tracking Food-Borne Illness the Crowd-Sourced Way 16

Posted by timothy
from the persistent-deliciousness-tracked-separately dept.
First time accepted submitter kraken9 writes "New research shows that online chatter can help you avoid food poisoning. Leveraging a statistical language model of Twitter users' online communication, nEmesis finds individuals who are likely suffering from a foodborne illness. People's visits to restaurants are modeled by matching GPS data embedded in the messages with restaurant addresses. As a result, each venue is assigned a health score based on the proportion of customers that fell ill shortly after visiting it. The paper shows that this score correlates with the official inspection data from the Department of Health, and argues that 'nEmesis offers an inexpensive way to enhance current methods to monitor food safety (e.g., adaptive inspections) and identify potentially problematic venues in near-real time.' Similar techniques have been used before to predict the spread of flu from GPS-tagged social data."
This discussion has been archived. No new comments can be posted.

There Is a Fly In My Tweet: Tracking Food-Borne Illness the Crowd-Sourced Way

Comments Filter:
    • by Nrrqshrr (1879148)
      Are you one of those lucky persons without a friend who feels obliged to tweet every plate she eats?
  • Any organism, virus, germ, bacterium that tries to get onboard me does so at it's own risk. I've got enough chemicals and heavy metals in my system to open my own compounding center, so bring it.

    But yes, I'm all for crowd sourcing on this one. Those little buggers might know how to travel, but they can't be the speed of the internet.
    • Those little buggers might know how to travel, but they can't beat the speed of the internet.

      Evolution is an endless arms race - tachyon based pathogens will evolve.

  • So all you have to do is spoof some GPS info on a Tweet (super easy to do), and then later be "sick" and you can take down a restaurant across the country? Bonus points for doing this across multiple accounts and making it appear like the shop had a bad case of food poisoning.

    It's interesting for tracking things like epidemics, but trying to use this to promote/tank business is going to turn into Yelp style shenannigans with the botnet spoofers making bundles of cash for rigging ratings.

  • How will they account for the widely varying incubation periods []of different foodborne illnesses?

    Two of the most common infections, E Coli and Salmonella, can take several days to show symptoms.

  • Assume people eat out, mention a restaurant in a tweet, and complain of being sick within 72 hours. I can see how over enough tweets -- just like enough searches -- you can make this work.

  • Hey! This could catch on.. Since the FDA has been bought, we might have a way of circumventing it.

It is much easier to suggest solutions when you know nothing about the problem.