Harnessing AI for Climate Change Communication: Analyzing Public Perception through NLP and Machine Learning

This study explores how Artificial Intelligence (AI), particularly Natural Language Processing (NLP) and machine learning, can be leveraged to analyze public discourse on climate change and improve climate communication strategies. Focusing on four key questions, the research examines how AI classifies public sentiment, identifies dominant topics, detects framing structures, and generates actionable insights to inform targeted communication. Over one million climate-related posts were collected from Twitter and Facebook between January and June 2024. Sentiment analysis using a fine-tuned BERT model categorized posts into positive, negative, and neutral sentiments, while Latent Dirichlet Allocation (LDA) revealed key topics. Framing analysis employed supervised machine learning to classify posts into narrative frames, and AI-powered visualization tools were used for interpretation. The results indicate a polarized sentiment distribution: 45% negative, 35% positive, and 20% neutral. Negative posts centered on skepticism and political inaction, while positive posts supported renewable energy and activism. Thematic analysis highlighted five key topics: scientific evidence, activism, economic implications, political debate, and environmental justice. Framing analysis revealed four dominant frames—climate urgency, economic impact, political action, and environmental justice—that shape public perception. Temporal sentiment shifts aligned with major events, suggesting public discourse is responsive to political and activist developments. This research demonstrates the potential of AI to provide scalable, data-driven insights into public climate discourse. By integrating these insights into strategic planning, communicators can design more effective, emotionally resonant messages, enhancing public engagement and supporting collective climate action.