Sentiment Analysis about COVID-19 vaccine on Twitter data: Understanding Public Opinion

Abstract

The COVID-19 pandemic has affected almost every segment of the worldwide population. Many vaccines were developed to cope with the ongoing COVID-19 outbreak. Indian government had initiated mass vaccination in early 2021 with objective of 100% vaccinated country. However, in spite of large number of awareness programs on the vaccination for the general public, still after more than one year 100% vaccination is not achieved. The reason for this is the hesitancy and lack of belief on the vaccination. The vaccination for children is now available and open for the general public. In order to boost the vaccination task, it is essential to understand the public opinion about the different vaccines available for adults and children. Social media platform, Twitter, is the most effective medium to know about the public perception and opinion about the vaccines. Therefore, an automated analysis of the public opinion using Twitter data is necessary to understand the mental thought process of public for the vaccine. The analysis allows the Indian governments to take some proactive measures to increase the public belief in the vaccine. Thus, the present study is intended to analyze the public sentiment on COVID-19 vaccination. 6000 real time tweets were fetched from the Twitter API with hashtag covid, covid19, vaccine, covaxin, covishield and childrenvaccine. The sentiments were analyzed and categorized into positive, negative and neutral sentiment by using TextBlob, Flair NLP, Stanza, and NLTK Vader algorithms. The performance of different algorithms was compared and evaluated. It was observed that NLTK Vader analyzed the public sentiment most correctly then the other algorithms. It was found that 34.210% sentiment were still negative about the vaccine which is a great matter of concern for the Indian governments.