Determining the Biases of Highly Trafficked,Neutral-Named Subreddits on Reddit

Abstract

This paper details the process behind evaluating the political biases of highly trafficked, neutrally named Subreddits on Reddit.com. Data used in this project is gathered from multiple Subreddits across Reddit. By analyzing the data within these posts and using techniques such as keyword analysis, domain analysis, sentiment analysis, and combining these metrics to define bias, one can discover the implicit bias within these Subreddits and determine whether they are truly neutral, as their names describe them to be. Source code for the project is available at: https://github.com/alexdipz/SeniorThesis.