Manny
|
|
Manny (for each letter of the team members) is an iOS App that uses Natural Language Processing from Microsoft Azure services to provide a personalized text suggestion.
The App works by analyzing an incoming text and extracting the main information containing words. For example: Incoming: "When is the party tonight?" Response: "Tomorrow at 8 pm". In this case the main information is the adverb, "when"; the noun, "party"; and the time, "tonight". The intent would get stored in the Azure Cloud as well as the response sent. Upon a new incoming text, the algorithm will check for similarities and if it matches an existent question, it will fill in the text field with the previously-sent answer. If the incoming question doesn't match any of the existent intents, it will store it as a new one together with the answer. We worked on this project for the Global AI Hackathon in San Diego 2017. There were three challenges. First, to use Microsoft Azure services to build an application that would help time-management and optimization. Second, to build a fake news detector application. Third, to integrate emotion detection algorithms into an application for improving human interaction through social media. We participated in the first challenge and were second. |
#Microsoft Azure
#AI
#Python
#iOS
|