Scholarlink

<p>This study examines U.S. public sentiment toward solar energy from 2013 to 2022 by analyzing 8 million social media posts using RoBERTa, a transformer-based deep learning algorithm. While sentiment has been generally positive, it has declined since 2016, driven by increasing negativity. The analysis reveals significant and widening regional disparities, with Republican-leaning and Southern U.S. regions showing more pronounced negativity. A two-way fixed-effects panel analysis at the municipality level indicates that positive sentiment is associated with high solar radiation and a larger remote-working population, while negative sentiment is linked to a greater prevalence of multifamily housing. These results highlight the complex interplay of environmental, socioeconomic, political, and technological factors in shaping solar energy sentiment and underscore the need for tailored, region-specific strategies. The computational approach, which combines natural language processing and geospatial analysis of social media data, provides a scalable framework for sentiment analysis across diverse topics and regions, extending beyond the specific case of solar energy in the United States.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Sentiment analysis of solar energy in U.S. Cities: a 10-year analysis using transformer-based deep learning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年2月22日

查看全文
  • Serena Y. Kim,
  • Crystal Soderman,
  • Lan Sang

摘要

This study examines U.S. public sentiment toward solar energy from 2013 to 2022 by analyzing 8 million social media posts using RoBERTa, a transformer-based deep learning algorithm. While sentiment has been generally positive, it has declined since 2016, driven by increasing negativity. The analysis reveals significant and widening regional disparities, with Republican-leaning and Southern U.S. regions showing more pronounced negativity. A two-way fixed-effects panel analysis at the municipality level indicates that positive sentiment is associated with high solar radiation and a larger remote-working population, while negative sentiment is linked to a greater prevalence of multifamily housing. These results highlight the complex interplay of environmental, socioeconomic, political, and technological factors in shaping solar energy sentiment and underscore the need for tailored, region-specific strategies. The computational approach, which combines natural language processing and geospatial analysis of social media data, provides a scalable framework for sentiment analysis across diverse topics and regions, extending beyond the specific case of solar energy in the United States.