Lexicon-Based Sentiment Analysis and Emotion Classification of Climate Change Related Tweets

Abstract

The concerns for a potential future climate jeopardy has steered actions by youths globally to call the governments to immediately address challenges relating to climate change. In this paper, using natural language processing techniques in data science domain, we analyzed twitter micro-blogging streams to detect emotions and sentiments that surround the Global youth Climate Protest (GloClimePro) with respect to #ThisIsZeroHour, #ClimateJustice and #WeDontHaveTime hashtags. The analysis follows tweet scrapping, cleaning and preprocessing, extraction of GloClimePro-related items, sentiment analysis, emotion classification, and visualization,. The results obtained reveal that most people expressed joy, anticipation and trust emotions in the #ThisIsZeroHour and #ClimateJustice action than the few who expressed disgust, sadness and surprise. #ClimateJustice conveys the most positive sentiments, followed by #ThisIsZeroHour and the #WeDontHaveTime .In all the evaluations, a considerable number of people express fear in the climate action and consequences. Thus, climate change stakeholders and policy makers should consider the sentiments and emotions expressed by people and incorporate such outcomes in their various programmes toward addressing the climate change challenges especially as it affects the ecosystem.