Younghun Jeong

Abu Dhabi, United Arab Emirates

December 11, 2022 · EMNLP 2022

Younghun is an AI Research Engineer at NAVER Search, where he builds production LLM systems for search and commerce. His work spans retrieval-augmented generation, structured query planning, grounded generation, agentic workflows, evaluation, model distillation, and large-scale inference. He has contributed to the launches of Cue:, AI Briefing, and Shopping Agent.

His recent research includes Tree-of-Question, a structured retrieval framework for complex Korean question answering presented at the NAACL 2024 Industry Track. Before joining NAVER, he worked at the KAIST Users & Information Lab with Professor Alice Oh on responsible NLP and led the construction of the Korean Offensive Language Dataset.

He received his M.S. in Computer Science from Sogang University under the supervision of Professor Jungyun Seo. He also holds a B.A. in Chinese Culture and a B.S. in Computer Science from Sogang University. His interests include reliable LLM applications, retrieval and agent systems, human–AI interaction, and responsible AI.

news

Mar 25, 2026	NAVER Shopping Agent is now live for conversational product discovery and shopping assistance—a launch I worked on. 🚀
Mar 23, 2025	I helped launch NAVER’s AI Briefing, bringing generative answers directly into search results. 🚀
Jun 16, 2024	I attended NAACL 2024 in Mexico City (June 16–21, 2024) and presented Tree-of-Question! 🇲🇽
Sep 20, 2023	NAVER Search launched Cue:, an LLM-powered conversational search experience I helped build. 🚀
Dec 19, 2022	I started to work fulltime at NAVER Search 👨 ‍💻

selected publications

NAACL

Tree-of-Question: Structured Retrieval Framework for Korean Question Answering Systems

Dongyub Lee, Younghun Jeong, Hwa-Yeon Kim, Hongyeon Yu, and 6 more authors

In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 6: Industry Track), Jun 2024

Abs Paper

We introduce Korean language-specific RAG-based QA systems, primarily through the innovative Tree-of-Question (ToQ) methodology and enhanced query generation techniques. We address the complex, multi-hop nature of real-world questions by effectively integrating advanced LLMs with nuanced query planning. Our comprehensive evaluations, including a newly created Korean multi-hop QA dataset, demonstrate our method’s ability to elevate response validity and accuracy, especially in deeper levels of reasoning. This paper not only showcases significant progress in handling the intricacies of Korean linguistic structures but also sets a new standard in the development of context-aware and linguistically sophisticated QA systems.
EMNLP

KOLD: Korean Offensive Language Dataset

Younghun Jeong, Juhyun Oh, Jongwon Lee, Jaimeen Ahn, and 3 more authors

In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, Dec 2022

Abs Paper Data

Recent directions for offensive language detection are hierarchical modeling, identifying the type and the target of offensive language, and interpretability with offensive span annotation and prediction. These improvements are focused on English and do not transfer well to other languages because of cultural and linguistic differences. In this paper, we present the Korean Offensive Language Dataset (KOLD) comprising 40,429 comments, which are annotated hierarchically with the type and the target of offensive language, accompanied by annotations of the corresponding text spans. We collect the comments from NAVER news and YouTube platform and provide the titles of the articles and videos as the context information for the annotation process. We use these annotated comments as training data for Korean BERT and RoBERTa models and find that they are effective at offensiveness detection, target classification, and target span detection while having room for improvement for target group classification and offensive span detection. We discover that the target group distribution differs drastically from the existing English datasets, and observe that providing the context information improves the model performance in offensiveness detection (+0.3), target classification (+1.5), and target group classification (+13.1). We publicly release the dataset and baseline models.