AI-Driven Social Media Analytics for Assessing Climate Change Perceptions and Supporting Adaptation and Sustainability Policies

This study examines public perceptions and discourse on climate change using artificial intelligence (AI)-based analysis of social media data, with implications for climate adaptation and sustainability policy. A dataset of 29,576 posts from the X platform (December 2025) was analyzed through an integrated framework combining text mining, TF-IDF-based word analysis, deep learning-based sentiment analysis, and Latent Dirichlet Allocation (LDA) topic modelling. The findings reveal that climate change discourse is predominantly characterized by negative sentiment, reflecting high levels of concern, perceived risk, and urgency, while positive content emphasizes awareness, solutions, and collective action. Topic modelling identifies three main themes: skepticism shaped by daily weather experiences, scientific and policy-oriented climate debates, and discussions on carbon emissions and human impact. These results demonstrate that social media serves not only as a space for emotional expression but also as a dynamic platform for information exchange and public opinion formation. From an adaptation perspective, AI-driven social media analytics provide valuable insights into public risk perception, misinformation patterns, and knowledge gaps, supporting evidence-based climate communication and policy development.