Moving from the delicious brunch, noodle, chicken, taco, and Mediterranean restaurants in Chicago's Lakeview neighborhood to the sad fast food reality of Chicago's Loop was somewhat of a shock to the employees at Datascope when we moved into a bigger office in 2012. Lucky for us, dozens of unique food trucks roam the streets of the Loop- if you can find them.
Rather than have to skim through Twitter timelines to figure out which trucks would be nearby on a given day, we built a web app to do the heavy lifting for us. lunch.datascopeanalytics.com (also known as Lunch?) reads in tweets from nearly 90 food trucks every few minutes, parses out locations like street corners and skyscrapers in the Loop and at the University of Chicago's Hyde Park campus, and locates the trucks on a map to easily show the user who's nearby. People with favorite trucks can also click their favorites in the sidebar to zap to the truck's most recent location and see its recent tweets.
Rather than try to build out highly complex NLP logic to ensure our app could parse schedules, times, misspelled street names, and other corner cases, we chose to make the actual Twitter data highly visible to the user– at a glance, they can see the tweets we used to infer a location and confirm for themselves whether the truck is still at that location.
The app is live at http://lunch.datascopeanalytics.com/.