Shopping Knowledge Base Construction

• Building a shopping knowledge base that extracts and refines contextual knowledge from UGC (blogs, reviews) and matches it to products, targeting queries where lexical matching fails to surface relevant items (e.g., “putter for beginners”, “housewarming gift”)

• May 2026

Shopping Guide Agents

• Developed LLM agents that provide shopping guides, suggestion chips, and product recommendations on search result pages and in conversational mode; launched as NAVER Shopping Agent (Beta) in Feb 2026

• Jun 2025 ~ May 2026

AI Purchase Guide: Query-Theme Modeling

• Expanded AI purchase guide coverage beyond the home-appliance category by shifting from per-category to per-query theme extraction with 2-step prompting

• 2025

C2C Search Query Understanding (Wallapop)

• Redesigned distance-based ranking (DMP) with shipping-availability-aware documents

• Maintained query term importance (QTI) models and Query Reformulation (QR) system

• Developed a term-level typo correction pipeline

• Dec 2023 ~ Dec 2024

C2C Search Query Understanding (Poshmark)

• Introduced algorithmic typo correction covering unseen queries

• Extended query reformulation from1:1 to 1:N using query-embedding-based pair generation, substantially expanding retrieved items per query

• Maintained query term importance (QTI) models and related-query suggestion

• Dec 2023 ~ Dec 2024

Clustering-based Image-Text Graph Matching for Domain Generalization

• Nokyung Park, Daewon Chae, Jeong Yong Shim, Sangpil Kim, Eun-Sol Kim, Jinkyu Kim

• International Conference on Pattern Recognition (ICPR) 2024

[Paper] [Code]

EGTR: Extracting Graph from Transformer for Scene Graph GenerationEGTR: Extracting Graph from
Transformer for Scene Graph Generation

• Jinbae Im, JeongYeon Nam, Nokyung Park, Hyungmin Lee, Seunghyun Park

• Computer Vision and Pattern Recognition Conference (CVPR) 2024 Oral

[Paper] [Code]

InstructBooth: Instruction-following Personalized Text-to-Image Generation

• Daewon Chae, Nokyung Park, Jinkyu Kim, Kimin Lee

• Preprint version

[Paper] [Project Page]

Bridging the Domain Gap by Clustering-based Image-Text Graph Matching

• Nokyung Park, Daewon Chae, Jeong Yong Shim, Sangpil Kim, Eun-Sol Kim, Jinkyu Kim

• ICML Workshop on  Spurious Correlations, Invariance, and Stability (SCIS) 2023

[Paper]

Grounding Visual Representations with Texts for Domain Generalization

• Seonwoo Min, Nokyung Park, Siwon Kim, Seunghyun Park, Jinkyu Kim

• European Conference on Computer Vision (ECCV) 2022

[Paper] [Code]