An interactive visualization tool to explore realtime tweets in New York City. Project developed for Large Scale Visual Analytics course by Enrico Bertini at NYU.
The server is a Python application running on Docker, and the live version is deployed on Google Cloud Engine. It uses tweepy to listen to geo-tagged real-time tweets from New York City. Tweets are preprocessed by NLP package spaCy for lemmatization and TextBlob for sentiment analysis, before being sent to the frontend via websockets.
The frontend is a static site built with D3js for visualization and interactions, plus crossfilter to support fast filtering with coordinated views. MapboxGL is used to visualize tweets on a map in combination with D3js for interaction.