Image of tool
This GitHub repository stores the code that was written to perform a sentiment anaylsis of tweets about the 2016 presidential candidates. This project was created during the Information Visualization course taught at NYU-Poly by Professor Bertini.
The project works as such: tweets were saved for a month long period and then a dictionary machine language approach was utilized to assign a sentiment to each tweet. These scores were then aggregated by state and candidate and subsequently visualized.
The screencast can be found at: https://vimeo.com/167437613
The documentation can be found at: https://docs.google.com/document/d/15LM7FoIhTKKL7BLVSj2K4YtfccN4yWH2cfj9-jfJNYU/edit?usp=sharing
The visualization can be found at: http://tweet-sentiment-nyu.herokuapp.com/ and no software needs to be installed to load this visualization.
The html files are in public folder, namely test.html and dorlingCartogram.html.