arXiv cs.AI· Hongwei Zhang, Qiqiang Zhong, Junfeng Shu, Jiangxia Cao, Yiyang Lv, Huanjie Wang, Liwei Guan, Jing Yao, Yiyu Wang, Zhaojie Liu, Han Li·· 6 天前AI 评分20
UxSID：面向超长序列的语义感知用户兴趣建模框架
UxSID: Semantic-Aware User Interests Modeling for Ultra-Long Sequence
AI 导读
UxSID 提出语义组共享兴趣记忆的第三条路径，利用 Semantic IDs（SIDs）与双层注意力策略，在无需 item-specific 模型高开销的情况下捕捉目标感知偏好。该端到端架构兼顾计算精简与语义感知，取得 SOTA 性能，并在大规模广告 A/B 测试中实现 0.337% 的收入提升。
正文 · 原文
Authors:Hongwei Zhang, Qiqiang Zhong, Junfeng Shu, Jiangxia Cao, Yiyang Lv, Huanjie Wang, Liwei Guan, Jing Yao, Yiyu Wang, Zhaojie Liu, Han Li
Abstract:Modeling ultra-long user sequences involves a difficult trade-off between efficiency and effectiveness. While current paradigms rely on either item-specific search or item-agnostic compression, we propose UxSID, a framework exploring a third path: semantic-group shared interest memory. By utilizing Semantic IDs (SIDs) and a dual-level attention strategy, UxSID captures target-aware preferences without the heavy cost of item-specific models. This end-to-end architecture balances computational parsimony with semantic awareness, achieving state-of-the-art performance and a 0.337% revenue lift in large-scale advertising A/B test.
| Comments: | Work in progress |
| Subjects: | Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Machine Learning (cs.LG) |
| Cite as: | arXiv:2605.09040 [cs.AI] |
| (or arXiv:2605.09040v4 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2605.09040 arXiv-issued DOI via DataCite |
Submission history
From: Hongwei Zhang [view email]
[v1]
Sat, 9 May 2026 16:26:48 UTC (2,829 KB)
[v2]
Wed, 13 May 2026 07:20:14 UTC (2,828 KB)
[v3]
Mon, 18 May 2026 07:00:17 UTC (2,828 KB)
[v4]
Fri, 9 Oct 2026 08:29:36 UTC (2,829 KB)
来源：arXiv cs.AI · arxiv.org