As someone that has been dealing with a particularly nasty bout of food poisoning over the last week or so, this is a story that is currently near and dear to my stomach.
While I cannot, and will not, publicly speculate as to which restaurant is responsible for my current state of misery, I can say this – if there is a voice in your head telling that the calamari at the bar you just ordered looks a little suspect….listen to it. It is probably right.
But pretty soon, I may have been able to avoid this situation using the power of social media.
Researchers at the University of Rochester have come up with a way to leverage the massive amounts of data online, specifically from Twitter, to help identify likely sources of food borne illness. Named “nEmesis”, this program is essentially a passive monitoring program that “listens” and combines the data from public tweets, geo-tagged Twitter messages, interprets the tweets to look for indications of food borne illness, and tracks user activity following a restaurant visit. Relevant hashtags like #tummyhurts or tweets where the user describes any sort of GI issues (“OMG just puked my face off”) gets flagged and that users relevant data is then analyzed. From that dataset, nEmesis then ranks the likelihood that a restaurant is the source of a food poisoning outbreak.
Sifting through 3.8 million tweets from 94,000 Twitter accounts in the New York area, nEmesis identified 480 instances of food poisoning. The system then pinpoints the restaurants that these people have visited and looks for commonalities and likely culprits.
While the system isn’t foolproof, and really only provides what amounts to estimates instead of hard evidence, their findings matched up closely with the New York Department of Health’s current grades for the identified establishments.
While the system isn’t perfected yet, and nEmesis has yet to be turned into a publicly available app yet, it could prove to be a useful tool for identifying sources of food borne illnesses.
- By Jon Watson, Food & More blog