Portraits of O Hwang Kwon, Katie Vu, Naman Bhargava, Mohammed I. Radaideh, Jacob Cooper, and Veda Joynt

AI-powered paper on nuclear sentiment wins NEUP award

NERS students are among the authors of the winning research paper.

A paper co-authored by several U-M Nuclear Engineering and Radiological Sciences (NERS) students earned top honors in the Undergraduate Competition of the 2025 Innovations in Nuclear Energy Research and Development Student Competition, a program supported by the Department of Energy’s Nuclear Energy University Program (NEUP).

NERS students O Hwang Kwon, Jacob Cooper, and Veda Joynt contributed to the interdisciplinary team alongside Computer Science undergraduate student Katie Vu, Master’s in Data Science student Naman Bhargava, Mechanical Engineering Postdoctoral Fellow Mohammed I. Radaideh, and NERS assistant professor Majdi Radaideh.

Since contributing to the project, several team members have taken exciting next steps. O Hwang completed his Master’s in May 2024 and has returned to the Korean military as an Army Captain. He will join a faculty position in the Mechanical Engineering department of Korean Westpoint. Jacob finished his undergraduate degree in May 2025 and is now working at Westinghouse. Veda graduated with her NERS undergraduate degree in May 2024 and is currently pursuing a master’s in data science at Leuphana University of Lüneburg in Germany. 

Naman earned his master’s in data science in May 2025 and is now working at Lumos Learning. Mohammed continues his work as a MICDE Research Scholar and postdoctoral fellow in the Artificial Intelligence & Multiphysics Simulations (AIMS) Lab

Katie, a rising computer science senior, has presented her research on natural language processing and large language models at multiple conferences, including the E-HAIL Symposium and the 2025 ASEE Annual Conference & Exposition. She is currently completing a research internship at Carnegie Mellon University and plans to pursue graduate studies in the same field. As a leading author on the paper, she was nominated to represent the team for the competition. The award recognizes the contributions of the entire student team, whose collaborative efforts brought together nuclear engineering, computer science, and data science to produce this impactful research.

The winning paper, Sentiment analysis of the United States public support of nuclear power on social media using large language models, explores how Americans feel about nuclear energy using artificial intelligence tools to analyze more than 1.26 million posts on X (formerly Twitter), collected between 2008 and 2023. A subset of 300,000 geotagged posts was annotated using large language models to evaluate sentiment across all 50 states. 

The team found that 48 states showed more positive than negative sentiment, with a national average of 54% positivity. Positive posts highlighted nuclear’s reliability, clean energy potential, and technological innovation, while negative posts commonly focused on waste, safety, and cost—offering insights into the opportunities and challenges facing nuclear energy adoption.

In addition to mapping public opinion, the team developed a novel, bias-reducing approach to sentiment classification. By using seven labeling tools and determining final sentiment through majority vote, they created a high-confidence training dataset that boosted model accuracy to 96%. 

This methodology not only strengthens the reliability of AI-driven sentiment analysis but also lays the groundwork for future tools—such as a real-time dashboard to monitor public opinion on nuclear energy across social media and news platforms. Sponsored by the Fastest Path to Zero Initiative and the Department of Energy, the project reflects the growing importance of interdisciplinary research in building an informed clean energy transition.

“This was the first funded project I had in Michigan through Fastest Path to Zero Initiative, and this exceptional computing team delivered remarkable work with amazing collaboration,” said Majdi Radaideh. “Their efforts not only created a strong paper but also paved the way for future grants and broader impact within our lab. I’m incredibly proud of each one of them.”

The paper was also highlighted in Michigan News. Pictured: O Hwang Kwon, Katie Vu, Naman Bhargava, Mohammed I. Radaideh, Jacob Cooper, and Veda Joynt.