Today I participated in a great panel discussion about Twitter and public health.
Here’s the view from the stage:
Here’s a description from the Social Media Week website:
People Tweet about medical problems 4 days before they see a doctor.
Most don’t know they can report food poisoning to the city.
Most food poisoning cases go unreported.
This is what drove the Smart Chicago Collaborative, in partnership with local civic developers Cory Nissen, Joe Olson, and Scott Robbin, and the Chicago Department of Public Health (CDPH), to launch Foodborne Chicago, an innovative application that scans Twitter for mentions of food poisoning in Chicago.
This web app enabled them to connect with Chicago residents on Twitter through@foodbornechi and encourage them to report details of their food poisoning to the Food Protection Division of CDPH. Through Online Open 311, residents also get to see the inspection results of their report through 311 Service Tracker Chicago which are listed in detail at theChicago Open Data Portal.
In order to determine if a tweet is relevant to “food poisoning” or just noise, the web app has to classify tweets by sifting through Chicago’s 50,000 tweets/day. This classification is done via machine learning using a mathematical, natural language model built on using existing tweets and human classifiers (Dan O’Neil and Raed Mansour).
Basically, the app learns which tweets from Chicago residents may have food poisoning, presents them to human classifiers, who @replies to these resident tweets to submit an online report to food inspectors to initiate an investigation.
Join our panelists below, with moderator Ankur Thakkar, Digital Director for the City of Chicago Mayor’s Office, as they discuss and answer questions about the web app, the future of using Twitter and Machine Learning and what they learned from communicating on Twitter to Improve Public Health.
Check-out the latest write-ups from the Chicago Tribune and Zagat. Also, for those wanting some deeper insight into the programming, read about the R Statistics Language and somebehind-the-scene programming used in FoodBorneChi.
Event Room: Claudia Cassidy Theater at the Chicago Cultural Center
Here’s video: