Visualizing Urban vs. Rural Sentiments in Real-Time

Abstract

Discrepancies in sentiment between urban and rural communities represent a divide which has garnered much media attention yet so far has yielded little research or analysis. In this research, we use sentiment analysis to parse tweets in order to reveal the mood of each demographic group when discussing specific topics. We expose this method through a publicly accessible web application for sentiment tracking. Users are able to track specific keywords on Twitter in order to collect data at different scales, filtering by country, state, or even neighborhood. Using this tool, we find that across a broad range of topics generally believed to be polarizing, urban and rural groups actually express very similar sentiment scores. These results suggest that even though two demographic groups might hold completely opposite views on an issue, there is usually a certain symmetry in the emotion that both groups bring to the discourse.