imhungry, is an Alexa Skill that recommends a restaurant based on the user's current sentiment levels. After activating the skill by saying "Hey Alexa, tell I'm hungry ...", the Alexa will begin parsing the user's speech and convert it into plain text. This plain text is then posted to IBM Watson's Sentiment Analysis API, and a sentiment score is returned. Based on the sentiment score, a set of preset descriptions of the user's preferred restaurant will be used - an extremely low sentiment score will recommend a cheap ice-cream parlor, in a close vicinity. These descriptions, along with the users location, will be posted to the Yelp Business Search API which returns a restaurant that best suites the user's description. This restaurant's longitude and latitude, and then placed into a cloud database service - Google Firebase. Finally, a local server is created which grabs the location from our created Firebase database, which produces a webpage that includes the chosen restaurant on a map using Google's Maps API (see videos in this repo for a demonstration of imhungry - RainyDay.mp4 and GoodExam.mp4).
This project was made during 36 hours at Hack the Valley 2 in a team of four - Arjun Bhushan (Software Engineering UW), Andy Jin (Computer Engineering UW), Sara Fong (Software Engineering UW), and Marcus Yung (Computer Engineering UW).