ENVIRONMENTAL SENTIMENT ANALYSIS: LEVERAGING AI TO ASSESS PUBLIC PERCEPTION OF ECOLOGICAL ISSUES THROUGH TEXT DATA FUSION

Abstract

Global warming, often referred to as climate change, is now emerging as one of among the most highly debated topics of over a decade or so. A lot of individuals think that warming temperatures pose a serious threat to our planet, even if some people claim it is a myth. This article examines how public opinions have changed over the last 10 years by using sentiment analysis to examine Twitter data. With 320 million active users each month, Twitter is a useful tool for determining public opinion. Using sentiment analysis, we extracted tweets that had terms like "global warming" and "climate change," classifying them according to whether they were neutral, positive, or negative. We trained numerous data sets utilizing Naïve Bayes, Multinomial Naïve Bayes equations and SVM-based classification algorithms for the purpose to reach highest possible accuracy. Then, employing data from Twitter, the approach with the highest accuracy rate has been employed to evaluate how perceptions on global warming have fluctuated over time.