It is important to understand stakeholders' attitudes toward energy transition policies, or indeed toward any government proposal. Online news platforms are agoras where media and people interact, so they appear to be a rich vein for mining media and public attitudes toward energy transition. We collected online news articles and comments with the keyword “energy transition,” along with such commenters' activity histories, from the largest Korean online news aggregator during the July 2021 to July 2022 shift in the government. Utilizing advanced machine learning approaches, we examined the media's evaluation of energy transition, users' emotions regarding it, and the influence of various features on commenters' emotions and elaborations in comments. We found that intense negative emotions dominated news comments about the energy transition. However, our analysis of users' historical activities found that the commenters were a raging political crowd who expressed intense negative emotions about many social and political issues. Furthermore, our application of Catboost regression analysis demonstrates how news framing can facilitate deliberation by relieving the intensity of emotions and enhancing users' deliberation. Our findings suggest being cautious about using online news platform data to gauge media and public attitudes toward energy transition. The findings also highlight the need to use comprehensive datasets and approaches to avoid misunderstanding stakeholders' attitudes toward energy policy, as expressed in online spaces. These findings can guide future studies that use news comments to understand public attitudes toward energy transition policies or other contentious proposals.

中文翻译：

---

转型之声：韩国在线新闻媒体和用户对能源转型的评论

了解利益相关者对能源转型政策或任何政府提案的态度非常重要。在线新闻平台是媒体和人们互动的场所，因此它们似乎是挖掘媒体和公众对能源转型态度的丰富脉络。我们从韩国最大的在线新闻聚合器收集了 2021 年 7 月至 2022 年 7 月政府换届期间关键词为“能源转型”的在线新闻文章和评论，以及此类评论者的活动历史记录。利用先进的机器学习方法，我们研究了媒体对能量转变的评价、用户对此的情绪，以及各种特征对评论者情绪和评论阐述的影响。我们发现，强烈的负面情绪主导了有关能源转型的新闻评论。然而，我们对用户历史活动的分析发现，评论者是一群愤怒的政治人群，他们对许多社会和政治问题表达了强烈的负面情绪。此外，我们对 Catboost 回归分析的应用展示了新闻框架如何通过缓解情绪强度和增强用户的深思熟虑来促进深思熟虑。我们的研究结果表明，使用在线新闻平台数据来衡量媒体和公众对能源转型的态度要谨慎。研究结果还强调需要使用全面的数据集和方法来避免误解利益相关者对在线空间中表达的能源政策的态度。这些发现可以指导未来的研究，利用新闻评论来了解公众对能源转型政策或其他有争议提案的态度。