Sentiment is a powerful motivation which reflects the ideology of politicians. Twitter has now become a common platform for the United States members of Congress to express and proliferate their agenda to the voter-base. Does Twitter play a role in the outcome for the Senate race elections in 2020? Is there a relationship between the sentiments of winners and losers on social media? In this paper, we analyze these key relationships between incumbent senators and their challengers from over thirty states, most of which are key battleground states. The Twitter feed for all the politicians is mined, cleaned and analyzed using sentiment analysis and statistical modelling using Python and MATLAB. We find that the difference between the rate of tweet and sentiment scores for sitting senators and their challengers from the Republican and Democratic party is significantly different. We also report that there is a notable difference between tweet rate and sentiment score for members within a party.

Using Twitter to predict the United States Senate Election Year 2020