Young-Soo Ko

Young Soo Ko is a PhD student in the College of Information Science at the University of Arizona. His research centers on natural language processing and large language models, with three interconnected threads: (1) computational text mining of social media and other user-generated content, (2) systematic evaluation of LLM capabilities and limitations on knowledge-intensive tasks and (3) scientometric analysis of how research communities produce and organize knowledge. His recent work includes a multidisciplinary benchmark showing that frontier LLMs systematically over-predict causality when reading scientific abstracts (Findings of EMNLP 2026) and an LLM-powered framework for hierarchical topic discovery published in Information Research (iConference 2026). Through this work, he hopes to advance NLP methods that are both rigorous and useful for real-world text analysis.