Twitter Could Be Used To Track Food Poisoning
In addition to sharing funny, pointless thoughts and helping to track disease, Twitter might be able to tackle the prevention of food poisoning next.
48 million people acquired some sort of foodborne illness in 2011 according to the Centers for Disease Control and a team of computer scientists from the University of Rochester believes that this is an entirely avoidable issue with the help of their new program, nEmesis.
nEmesis uses an algorithm that seeks out tweets suggesting the user is ill. The geocoordinates that are embedded in the message help to link the tweet to its location and crowdsourcing helps to catch tweets from users who might have foodborne illness. The unique program combines the efforts of both humans and computers. The study explains, "We are able to do this by leveraging vast sensor networks composed of users of online social media, who report -- explicitly as well as implicitly -- on their activities from their smart phones."
nEmesis "listens" to tweets that are tagged at specific restaurants and then continues to track the users for 72 hours. If the users report symptoms of illness via Twitter, nEmesis can follow it back to where the user had eaten.
Over a four-month period, nEmesis sorted through the data of 3.8 million tweets from 94,000 users in New York City. From those tweets, the system kept track of 23,000 restaurant-goers and found 480 probable reports of foodborne illness. The researchers then ranked the restaurants by their likelihood of passing on food poisoning to a customer. This ranking was analyzed in comparison with the inspection results of the New York City Department of Health. The researchers discovered an overlap of about a third between the two resources.
While these numbers demonstrate the nEmesis still has some work to do, Adam Sadilek, who created the program as a postdoctoral fellow at Rochester, still believes that the real time data of nEmesis can benefit everyone. New York City's Department of Health inspects restaurants only once a year. These unannounced visits can be better organized if nEmesis manages to alert officials to inspect a restaurant in accordance with reports of food poisoning. Sadilek hopes to collaborate with health agencies so that his technology can better inform the public with updated restaurant rankings.
This promising technology sounds like a much better option than a night spent feeling sick.
check out the slideshow below for the most common items that cause food poisoning.
Image Credit: nEmisis/University of Rochester