Jon Atkins

This is my final project for an Introduction to Geographical Information Systems (GIS) class I took at Tufts University during the Spring 2016 Semester.

The project, which took place over the course of about a week was an attempt to use geographic and text data contained in political tweets and historical election precinct-level election results to analyze trends in the discussion around US electoral politics.

My hypothesis was that places that are more geographically polarized (precincts measured as geogrpahic outliers based on past election results) were more likely to engage in more aggressive or negative political conversation online (measured as a sentiment analysis of politically-related tweets).

More importantly, the goal was to better familiarize myself with tools like the Twitter Search API (particularly with regards to geodata), Google Geocoder, Sentiment Analysis, ArcMAp, and precinct-level election data, as well as to work on my GIS and coding skills.

Please view the accompanying paper or final poster for a more elaborate discussion of the results.