Evaluating Sentiment Classification Techniques for Analyzing COVID-19 Vaccination Opinions on Twitter Data

This analysis examines public sentiment regarding the COVID-19 vaccination campaign in India through Twitter. Many people used online forums to vent their worries and feelings during the worldwide epidemic, which was marked by lockdowns and extensive emotional, financial, and psychological effects. The work focuses on analyzing Twitter opinions about the necessity, administration, and trustworthiness of the vaccine, specifically concerning its immediate certification. Both of Lexicon-based or machine learning methods were used to do sentiment analysis after tweets had been pre-processed to eliminate noise and unnecessary information. The Lexicon-based approach employed VADER and TextBlob libraries in Python, while machine learning utilized Random Forest and Support Vector Machine (SVM) algorithms with TF-IDF feature extraction. The dataset consisted of 25,000 tweets, divided into three categories: negative, neutral, and positive. The outcomes showed that most tweets expressed neutral or positive sentiments. With an accuracy of 81%, the Random Forest algorithm using TextBlob labelling outperformed SVM, which had an accuracy of 79%, among the approaches tested. This indicates that the Random Forest model, using TextBlob for sentiment labeling, is the most effective for sentiment analysis in this context.