Andre Gala-Garza
Research Mentor(s): Majdi Radaideh
Mentor Department: Nuclear Engineering and Radiological Sciences
Authors: Andre Gala-Garza, Majdi Radaideh, Mohammed Al-Radaideh
Session: Session 5 (2:00pm – 2:50pm)
Presentation Type: Poster 37
Abstract
Using large language models (LLMs) for sentiment analysis or automatically identifying opinions expressed in text is a relatively new and rapidly advancing field. Another field that has made such advancements is nuclear power, both in energy and political contexts. Despite the significant controversy generated by discussions of nuclear power, comparatively little research focuses on revealing public sentiment towards nuclear power on a large scale. This project focuses on the development of a real-time software pipeline that captures text related to nuclear power from a variety of social media platforms, analyzes the sentiment of these posts using a pre-trained LLM, and updates a dashboard to display in real time the public sentiment of nuclear power in the United States. The data used for this analysis was scraped from short-form posts and news articles on websites such as Twitter, Reddit, Bluesky, and the New York Times. Analysis of these texts revealed that most posters on social media are negative towards nuclear power, with only a tiny fraction of posters expressing positive sentiment. Condensing these findings into an online dashboard enables direct navigation of sentiments from different groups, which may lead to a significant increase in public engagement with nuclear energy. In turn, this interest may encourage the public to become more attentive to a transition from using fossil fuels for power generation to cleaner sources such as nuclear power.
