The adoption of solar energy is pivotal in addressing climate change and achieving long-term energy security. However, its widespread deployment faces notable barriers, including high upfront costs, consumer doubts about system reliability, and unclear policy landscapes. To explore public perception and barriers to adoption, this study proposes a hybrid sentiment analysis framework that integrates a fine-tuned BERT model with TF-IDF-based feature enhancement. The model is applied to diverse consumer-generated content sourced from social media, review platforms, and public forums. Because the corpus is mixed-access, social-platform data were collected via official APIs and are not redistributed as raw text; instead, we share only permitted identifiers for rehydration (where allowed), preprocessing scripts, and non-reversible derived artifacts, while open corpora are shared in accordance with their licenses. Our approach achieves a validation F1-score of 0.85 and an overall test F1-score of 0.82, accurately capturing nuanced sentiments across domains. Quantitative analysis reveals that cost-related concerns account for over 41% of negative sentiment, followed by reliability (28%) and environmental skepticism (19%). The inclusion of cumulative gain analysis and high-confidence prediction filtering improves result interpretability and prioritization of insights. These findings provide valuable guidance for policymakers, solar energy firms, and sustainability advocates seeking to design targeted interventions and accelerate consumer acceptance of solar energy technologies.
Jabbar et al. (2026) studied this question.