3 papers
cs.IR2026
Query Generation with Direct Preference Optimization for Document Expansion in E-commerce Search
Kaihao Li, Feng Liu, Juexin Lin +4
Doc2Query, a popular document expansion technique, leverages sequence-to-sequence models to generate relevant queries, effectively addressing the "vocabulary mismatch" problem in i…
cs.IR2026
Scaling and Stabilizing Large-Scale Embedding-Based Retrieval
Zhen Yang, Juexin Lin, Hongwei Shang +8
Embedding-based retrieval (EBR) is foundational to large-scale e-commerce search, yet its effectiveness is often constrained by the quality of training signals and the representati…
cs.IR2024
Doc2Token: Bridging Vocabulary Gap by Predicting Missing Tokens for E-commerce Search
Kaihao Li, Juexin Lin, Tony Lee
Addressing the "vocabulary mismatch" issue in information retrieval is a central challenge for e-commerce search engines, because product pages often miss important keywords that c…