Soyoung Park
ICML 2026
Conference
ConferenceQuantile-Free Uncertainty Quantification in Graph Neural Networks
International Conference on Machine Learning
Quantile-free prediction intervals for uncertainty quantification in graph neural networks.
Soyoung Park
Postdoctoral Researcher · InnoCORE · UNIST
I am a postdoctoral researcher at Ulsan National Institute of Science and Technology (UNIST), affiliated with the Safe AI Research Center under the InnoCORE program. I work with Prof. Yeon-Chang Lee in the Data Intelligence Lab.
My research focuses on trustworthy graph learning, particularly uncertainty quantification and fairness. I also explore LLM-guided recommendation, temporal modeling with dynamic graphs, and multimodal learning.
Selected Research
SIGIR 2026Conference
ConferenceDisCoRec: Disentangled Conformity-aware Recommendation with LLM-Guided Multi-View Learning
ACM SIGIR Conference on Research and Development in Information Retrieval
LLM-guided multi-view learning for disentangled, conformity-aware recommendation.
CIKM 2025Conference
ConferenceFnRGNN: Distribution-aware Fairness in Graph Neural Network
ACM International Conference on Information and Knowledge Management
Distribution-aware fairness in graph neural networks through interventions at the structure, representation, and prediction levels.
Latest News
- Joined UNIST as a postdoctoral researcher in the InnoCORE program.
- Received my Ph.D. in Artificial Intelligence from Chungnam National University.
- Presented in the International Conference Research Highlights session at CKAIA. Presentation details →
- Presented in the Top Conference session at KCC. Presentation details →
- Received the Outstanding Research Award from CNU and was recognized as an Outstanding Researcher at the CNU Research Fair.