Investigation of trend change in public perception of coronavirus vaccines over time using Twitter data with artificial intelligence-assisted sentiment analysis

As of today, social media data provides great opportunities and supports us to study on realistic sentiment analysis with real life data. One of the sentiment analysis is to put forward meaningful sentiment distinctions from many Tweets to explore how people feel about an issue or a topic. Covid-19 is a new disease that is not yet fully understood. People are divided about vaccination developed against this disease. In this study, an analysis was conducted on the trend of public perception over time by using Twitter messages published in English regarding Covid-19 vaccines. Tweets are generally short in length and may contain emoticons, ironies, allusions or various spelling mistakes in written messages. From this perspective, it is hardly requirement to process the data before the analysis in order to get accurate results from the data analysis. A model was developed in the Google Colab environment using natural language processing algorithms and Python software language. The data were maintained for the analysis with the help of this model. Later on, VADER sentiment analysis, time series analysis and BERT machine learning analysis were applied on these data. The results obtained under the same conditions totally and the outputs of the results are presented.