Transformer-Based Methods for Sentiment Analysis in Climate Change Conversations: A Comparative Review

Climate change poses a critical global threat, influencing the environment, economy, and human well-being. Key drivers like increasing greenhouse gas emissions, deforestation, and industrial activities are contributing to substantial environmental shifts. Public sentiment about climate change plays a pivotal role in shaping policies and advocating for actions to mitigate its impacts. Incorporating sensor data into sentiment analysis introduces a fresh approach to capturing real-time environmental conditions, providing an additional layer of insight to complement traditional text-based sentiment analysis. This research aims to address existing gaps by comparing the effectiveness of various transformer models for detecting climate change sentiment. Although sentiment analysis has predominantly relied on textual data, there is a noticeable gap in exploring the potential of sensor data in this area. Previous studies have focused primarily on analyzing social media posts, news reports, and other written sources, often overlooking the valuable perspective that sensor data can provide. This study seeks to bridge this gap by integrating environmental sensor data to enhance the accuracy and contextual relevance of climate change sentiment analysis [15]. The main goal is to evaluate the capabilities of different transformer models in analyzing climate change sentiment. This research will assess each model's strengths and limitations, determine the most effective one for specific elements of sentiment analysis, and offer recommendations to guide future research in this field. By increasing the variety of data sources and the overall precision of climate sentiment detection, the results seek to improve sentiment analysis approaches.