arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2511.01461cs.IR
收藏

CAT-ID$^2$:面向电商生成式检索的类别树集成文档标识符学习

CAT-ID$^2$: Category-Tree Integrated Document Identifier Learning for Generative Retrieval In E-commerce

Beihang University(北京航空航天大学);Chinese Academy of Science(中国科学院)

Xiaoyu Liu, Fuwei Zhang, Yiqing Wu, Xinyu Jia, Zenghua Xia, Fuzhen Zhuang, Zhao Zhang, Fei Jiang, Wei Lin

AI总结:

针对电商生成式检索中忽略类别信息的问题,提出CAT-ID²方法,通过层次类别约束、聚类规模约束和离散度损失三个模块将类别树信息融入文档标识符,使相似文档ID更相似且保持唯一性,在线实验显著提升了模糊和长尾查询的订单量。

摘要:

生成式检索(GR)作为一种整合了大语言模型(LLMs)能力的有效范式,受到了广泛关注。它通常包含两个阶段:为文档构建离散语义标识符(IDs),以及通过自回归生成ID词元来检索文档。GR中的核心挑战在于如何构建具有强表示能力的文档标识符(DocIDS)。好的标识符应具备两个关键属性:相似文档应具有更相似的ID,且每个文档应保持独立且唯一的ID。然而,现有大多数方法忽略了电商中普遍且关键的原始类别信息。因此,我们提出了一种新颖的标识符学习方法——类别树集成文档标识符(CAT-ID$^2$),将先验类别信息融入语义IDs中。CAT-ID$^2$包含三个关键模块:用于在量化过程中逐层整合类别信息的层次类别约束损失、用于实现ID词元均匀分布的聚类规模约束损失,以及用于提升重构文档区分度的离散度损失。这些组件使CAT-ID$^2$能够生成使相似文档更相似、同时保留不同文档表示唯一性的标识符。大量离线和在线实验证实了我们方法的有效性,在线A/B测试显示,模糊意图查询的每千用户平均订单量提升了0.33%,长尾查询提升了0.24%。

英文摘要:

Generative retrieval (GR) has gained significant attention as an effective paradigm that integrates the capabilities of large language models (LLMs). It generally consists of two stages: constructing discrete semantic identifiers (IDs) for documents and retrieving documents by autoregressively generating ID tokens. The core challenge in GR is how to construct document IDs (DocIDS) with strong representational power. Good IDs should exhibit two key properties: similar documents should have more similar IDs, and each document should maintain a distinct and unique ID. However, most existing methods ignore native category information, which is common and critical in E-commerce. Therefore, we propose a novel ID learning method, CAtegory-Tree Integrated Document IDentifier (CAT-ID$^2$), incorporating prior category information into the semantic IDs. CAT-ID$^2$ includes three key modules: a Hierarchical Class Constraint Loss to integrate category information layer by layer during quantization, a Cluster Scale Constraint Loss for uniform ID token distribution, and a Dispersion Loss to improve the distinction of reconstructed documents. These components enable CAT-ID$^2$ to generate IDs that make similar documents more alike while preserving the uniqueness of different documents' representations. Extensive offline and online experiments confirm the effectiveness of our method, with online A/B tests showing a 0.33% increase in average orders per thousand users for ambiguous intent queries and 0.24% for long-tail queries.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑