OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Chen, Ben, Guo, Xian, Wang, Siyuan, Liang, Zihan, Lv, Yue, Ma, Yufei, Xiao, Xinlong, Xue, Bowen, Zhang, Xuxin, Yang, Ying, Dai, Huangyu, Xu, Xing, Zhao, Tong, Peng, Mingcan, Zheng, Xiaoyang, Wang, Chao, Zhao, Qihang, Zhai, Zhixin, Zhao, Yang, Liu, Bochao, Lv, Jingshan, Liang, Xiao, Ding, Yuqing, Chen, Jing, Lei, Chenyi, Ou, Wenwu, Li, Han, Gai, Kun
Format:	Preprint
Published:	2025
Subjects:	Information Retrieval
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866909864017526784
author	Chen, Ben Guo, Xian Wang, Siyuan Liang, Zihan Lv, Yue Ma, Yufei Xiao, Xinlong Xue, Bowen Zhang, Xuxin Yang, Ying Dai, Huangyu Xu, Xing Zhao, Tong Peng, Mingcan Zheng, Xiaoyang Wang, Chao Zhao, Qihang Zhai, Zhixin Zhao, Yang Liu, Bochao Lv, Jingshan Liang, Xiao Ding, Yuqing Chen, Jing Lei, Chenyi Ou, Wenwu Li, Han Gai, Kun
author_facet	Chen, Ben Guo, Xian Wang, Siyuan Liang, Zihan Lv, Yue Ma, Yufei Xiao, Xinlong Xue, Bowen Zhang, Xuxin Yang, Ying Dai, Huangyu Xu, Xing Zhao, Tong Peng, Mingcan Zheng, Xiaoyang Wang, Chao Zhao, Qihang Zhai, Zhixin Zhao, Yang Liu, Bochao Lv, Jingshan Liang, Xiao Ding, Yuqing Chen, Jing Lei, Chenyi Ou, Wenwu Li, Han Gai, Kun
contents	Traditional e-commerce search systems employ multi-stage cascading architectures (MCA) that progressively filter items through recall, pre-ranking, and ranking stages. While effective at balancing computational efficiency with business conversion, these systems suffer from fragmented computation and optimization objective collisions across stages, which ultimately limit their performance ceiling. To address these, we propose \textbf{OneSearch}, the first industrial-deployed end-to-end generative framework for e-commerce search. This framework introduces three key innovations: (1) a Keyword-enhanced Hierarchical Quantization Encoding (KHQE) module, to preserve both hierarchical semantics and distinctive item attributes while maintaining strong query-item relevance constraints; (2) a multi-view user behavior sequence injection strategy that constructs behavior-driven user IDs and incorporates both explicit short-term and implicit long-term sequences to model user preferences comprehensively; and (3) a Preference-Aware Reward System (PARS) featuring multi-stage supervised fine-tuning and adaptive reward-weighted ranking to capture fine-grained user preferences. Extensive offline evaluations on large-scale industry datasets demonstrate OneSearch's superior performance for high-quality recall and ranking. The rigorous online A/B tests confirm its ability to enhance relevance in the same exposure position, achieving statistically significant improvements: +1.67% item CTR, +2.40% buyer, and +3.22% order volume. Furthermore, OneSearch reduces operational expenditure by 75.40% and improves Model FLOPs Utilization from 3.26% to 27.32%. The system has been successfully deployed across multiple search scenarios in Kuaishou, serving millions of users, generating tens of millions of PVs daily.
format	Preprint
id	arxiv_https___arxiv_org_abs_2509_03236
institution	arXiv
publishDate	2025
record_format	arxiv
spellingShingle	OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search Chen, Ben Guo, Xian Wang, Siyuan Liang, Zihan Lv, Yue Ma, Yufei Xiao, Xinlong Xue, Bowen Zhang, Xuxin Yang, Ying Dai, Huangyu Xu, Xing Zhao, Tong Peng, Mingcan Zheng, Xiaoyang Wang, Chao Zhao, Qihang Zhai, Zhixin Zhao, Yang Liu, Bochao Lv, Jingshan Liang, Xiao Ding, Yuqing Chen, Jing Lei, Chenyi Ou, Wenwu Li, Han Gai, Kun Information Retrieval Traditional e-commerce search systems employ multi-stage cascading architectures (MCA) that progressively filter items through recall, pre-ranking, and ranking stages. While effective at balancing computational efficiency with business conversion, these systems suffer from fragmented computation and optimization objective collisions across stages, which ultimately limit their performance ceiling. To address these, we propose \textbf{OneSearch}, the first industrial-deployed end-to-end generative framework for e-commerce search. This framework introduces three key innovations: (1) a Keyword-enhanced Hierarchical Quantization Encoding (KHQE) module, to preserve both hierarchical semantics and distinctive item attributes while maintaining strong query-item relevance constraints; (2) a multi-view user behavior sequence injection strategy that constructs behavior-driven user IDs and incorporates both explicit short-term and implicit long-term sequences to model user preferences comprehensively; and (3) a Preference-Aware Reward System (PARS) featuring multi-stage supervised fine-tuning and adaptive reward-weighted ranking to capture fine-grained user preferences. Extensive offline evaluations on large-scale industry datasets demonstrate OneSearch's superior performance for high-quality recall and ranking. The rigorous online A/B tests confirm its ability to enhance relevance in the same exposure position, achieving statistically significant improvements: +1.67% item CTR, +2.40% buyer, and +3.22% order volume. Furthermore, OneSearch reduces operational expenditure by 75.40% and improves Model FLOPs Utilization from 3.26% to 27.32%. The system has been successfully deployed across multiple search scenarios in Kuaishou, serving millions of users, generating tens of millions of PVs daily.
title	OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search
topic	Information Retrieval
url	https://arxiv.org/abs/2509.03236