In fact, one out of every six US residents gets food poisoning every year. And where do people go to complain when they get sick?
One company is looking to tap into this resource of stomach suffering by scanning Twitter to figure out what restaurants are getting their customers sick. It's called nEmesis and it's developed by computer-science researchers at the University of Rochester.
The program uses natural language processing and AI to filter through tweets and connect the tweets to local restaurants via geotagging to identify the poisoning hotspots. It looks for tweets related to food poisoning, such as "I feel nauseous" and #upsetstomach, etc.
Social media monitoring and data mining isn't a new concept. It's been mostly used by companies to see what consumers are saying about their brand -- both good and bad. This is a novel and social conscious thinking use of that concept with some extra computer programming thrown in.
The City of Las Vegas's health department has been testing the program and has included nEmesis results into it's decisions of which food establishments to inspect. nEmesis scanned through 16,000 tweets a day for three months to create a high-priority list for inspectors.
Analyzing the results of the experiment, researchers found their Tweet-based system led to health violation citations in 15% of inspections; the city's former system resulted in citations just 9% of the time. Some of the citations included shutting venues down. nEmesis also identified 11 restaurants that posed a serious health risk while their control group identified just seven.
The study and program yielded 9,000 fewer food poisoning and 557 fewer hospitalizations than Las Vegas's average. Not bad.
What do you think about cities mining social media to compile restaurant inspection lists? Dig it? Hate it? Let us know in the comments below.