WeSearch：微信搜索的统一模型中心检索框架

Authors:Kepu Zhang, Yuzhe Cai, Yue Shu, Dawei Chen, Jinhai Zhu, Yuan Xu, Zehua Dai, Tongke Ni, Ziyao Chen, Chen Fu, Jie Shao, Chao Wang, Xiangmin Wang, Ruifeng Zhao, Rongchun Yao, Jie Yin, Weixin Qiu, Nan Song, Junyang Li, Fan Li, Quanchao Hui, Yulong Ou, Qiang Chen, Hao Cao, Sen Yang

View PDF HTML (experimental)

Abstract:We present WeSearch, a unified, model-centric retrieval framework for large-scale, continuously evolving content search. Prevailing methods assign documents fixed hierarchical Semantic IDs, implicitly assuming each document follows a single coherent semantic path-an assumption that often breaks down for multi-topic documents. WeSearch instead retrieves query-relevant clusters and then ranks atomic documents within the resulting query-specific candidate space, decoupling candidate pruning from fine-grained relevance modeling. Combined with relevance-aware training, compact document representations, and incremental index updates, this design enables weekly-or daily when needed-model refreshes. Deployed as the primary semantic retrieval path in WeChat Search alongside a lexical syntax tree, WeSearch shows consistent one-week A/B gains on corpora from hundreds of millions to billions of documents, and a one-month full-traffic A/B test confirms stable improvements of +0.68% CTR and +2.27% per-user article clicks, together with a 41% reduction in retrieval resource cost and a 19 ms reduction in graphic-text retrieval latency. WeSearch is now fully deployed in WeChat Search, serving 100% of production traffic.

Subjects:	Information Retrieval (cs.IR); Computation and Language (cs.CL)
Cite as:	arXiv:2611.00530 [cs.IR]
	(or arXiv:2611.00530v1 [cs.IR] for this version)
	https://doi.org/10.48550/arXiv.2611.00530 arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Kepu Zhang [view email]
[v1] Sat, 31 Oct 2026 01:35:20 UTC (837 KB)