This study examines how public perceptions of nuclear energy in Turkey are shaped through digital environmental activism on social media. Addressing a gap in the literature, it offers one of the first empirical analyses explicitly focused on Turkey's nuclear energy debate and the role of hashtag-based activism within it. The empirical dataset consists of 1831 posts from Twitter (now X), including both user-generated content and messages shared by Greenpeace Turkey during key events related to the Akkuyu Nuclear Power Plant. Utilizing natural language processing and supervised machine learning methods, the study employs multilingual sentiment analysis, text classification, and geospatial mapping to examine the dynamics of online discourse.
The results show that negative and neutral sentiments dominate public reactions, particularly in the Mediterranean region around Mersin, Turkey, where concerns about environmental risks and safety are most prominent. Activist-framed messages receive higher engagement, highlighting the mobilizing effects of environmental justice narratives. Methodologically, the findings demonstrate that translation-based pipelines using English-language models outperform Turkish-only models, revealing persistent resource gaps for under-resourced languages. Overall, the study provides a computational perspective on how social media shapes Turkey's nuclear energy discourse and offers insights for more context-sensitive communication and energy governance strategies.
扫码关注我们
求助内容:
应助结果提醒方式:

