Sentiment analysis on USA presidential election 2020

Abstract

Sentimental analysis is a technique in text mining used to define the sentiments usually positive or negative from textual input programmatically. It can be implemented in various fields such as customer feedback, movie or product reviews, and political comments. Companies implement sentiment analysis on comments in social media to analyse public opinion, perform market research, monitor brand, product reputation and comprehend customer experiences. Moreover, for a politician, this can be a powerful tool to analyse how well their actions are being received by the public. This project explores the use of scraping tweets from Twitter that mentioned the candidates' President Donald Trump and former Vice President Joe Biden from the span of January 1st to September 14th, 2020 to investigate the public sentiments and determine their approval ratings. This project is still undergoing the automated deployment which will allow it to still monitor the candidates up until the election date. For this project, the author would be building three different classifiers to help calculate the approval ratings for the presidential candidates. The first two models would be classifying the sentiments of tweets that either one of the candidates is present in. While the last model would help classify a tweet that contains both candidates and identify who mainly it is referring to.