Sentiment analysis of global news on environmental issues: insights into public perception and its impact on low-carbon economy transition

Abstract

In this study, we leverage sentiment analysis to investigate public perception towards environmental issues as conveyed through global news articles and its potential implications on the transition to a low-carbon economy. Utilizing an extensive corpus of news articles sourced globally, we deploy Natural Language Processing (NLP) techniques to quantify sentiment in these articles, capturing public sentiment’s dynamism and complexity towards various environmental issues. Our methodology involves sentiment scoring of key aspects like “climate change”, “climate policy”, “renewable energy”, “solar energy”, “wind energy”, and “environmental impact” which facilitated a detailed sentiment trend analysis over time. We also incorporated a Latent Dirichlet Allocation (LDA) model to conduct topic modelling, identifying five major topics recurring in the discourse. Our correlation analysis uncovers interesting relationships such as a positive correlation between sentiment scores of “low carbon” and “electric cars”, and a negative correlation between “greenhouse gas emissions” and “electric cars". The findings indicate that public sentiment towards environmental issues is not only multifaceted but also evolving, with significant implications for policy-making and stakeholder engagement in the low-carbon transition. These results exemplify sentiment analysis as a powerful tool in understanding public perception, providing actionable insights for researchers, policymakers, and stakeholders involved in environmental issues and the low-carbon economy transition.