As avid followers of the ever-changing political battlefield, as the 2016 presidential election draws closer and closer, our team decided it would be a fun and interesting project to create a natural language processing-based internet parser that utilizes a system of ranking and sentiment analysis to generate a time-evolving map of the United States which tracks sentiment towards candidates by geographic region. This is encapsulated in the form of a web application to allow for ease of data visualization and provide accurate, up-to-date information for users who desire a quick analysis and prediction of election results.
Our project makes use of several dependencies, including the nltk and tweepy libraries. We have customized these packages to conform to the specifications of our implementation.