Machine Learning based Twitter Sentiment Analysis on COVID-19

Corona virus is a deadly disease that has been spreading around the world for many years. COVID-19 is a new variant of coronavirus which causes its least attack as cold, fever and headache to its severe attack as breathing problems and death. Twitter is a social network tool where a large number of people communicate and express their feelings through tweets and posts in everyday life. Our task is to analyze the perception of people from all over the world during this pandemic. So, analyzing the sentiment of COVID-19 tweets will be one way to find out how people's emotions and the situations are in the world. This system uses various vectorizers, NLP and NLTK tools, which can be used while building various machine learning models for n-grams, and, consequently, to classify tweets into bi-class and tri-class. This system shows a maximum accuracy of 96.7%.