aiqbench analysis
This research addresses a critical gap in e-commerce search by introducing DSIRM, a model that injects explicit query-item supervision into discrete semantic identifier generation to overcome the limitations of unsupervised quantization. By combining contrastive quantization on the item side with generative LLMs for query interpretation, the approach effectively resolves tail query ambiguity and enhances query-dependent ranking. The method demonstrates significant industrial viability through a hybrid architecture that delivers measurable offline AUC improvements and substantial online lifts in user click-through rates and conversion rates on production data.
DSIRM: Learning Query-Bridged Discrete Semantic Identifiers for E-commerce Relevance Modeling
arXiv:2606.04374v1 Announce Type: cross Abstract: Despite rapid progress of continuous embeddings for e-commerce search relevance, a long-standing open problem is the difficulty in capturing fine-grained attribute distinctions. While discrete Semantic Identifiers (SIDs) have been widely adopted as a promising alternative, existing SID generation methods rely heavily on unsupervised quantization. In realistic scenarios, the lack of explicit supervision often makes it more difficult to dictate which items should share an SID, resulting in limited capability for query-dependent ranking. To address the issue of unsupervised SIDs, we propose to explicitly model discrete relevance features and develop a Discrete Semantic Identifier Relevance Model (DSIRM). Specifically, we present a query-bridged contrastive quantization approach on the item side, injecting query-item interaction supervision into Residual Quantization to actively learn relevance-aware semantic partitions. On the other hand, we explore generative LLMs on the query side to explicitly predict item SIDs from text, resolving tail queries and intent ambiguity. Hierarchical prefix matching between query and item SIDs yields discriminative features that perfectly complement dense signals. Extensive experimental results on Tmall's production data show that our proposed approach has achieved better results, improving offline AUC by +1.54\%. Deployed via an efficient hybrid architecture, it achieves significant online lifts (+0.13\% UCTR, +0.25\% UCTCVR), proving its massive industrial value.
Abstract:Despite rapid progress of continuous embeddings for e-commerce search relevance, a long-standing open problem is the difficulty in capturing fine-grained attribute distinctions. While discrete Semantic Identifiers (SIDs) have been widely adopted as a promising alternative, existing SID generation methods rely heavily on unsupervised quantization. In realistic scenarios, the lack of explicit supervision often makes it more difficult to dictate which items should share an SID, resulting in limited capability for query-dependent ranking. To address the issue of unsupervised SIDs, we propose to explicitly model discrete relevance features and develop a Discrete Semantic Identifier Relevance Model (DSIRM). Specifically, we present a query-bridged contrastive quantization approach on the item side, injecting query-item interaction supervision into Residual Quantization to actively learn relevance-aware semantic partitions. On the other hand, we explore generative LLMs on the query side to explicitly predict item SIDs from text, resolving tail queries and intent ambiguity. Hierarchical prefix matching between query and item SIDs yields discriminative features that perfectly complement dense signals. Extensive experimental results on Tmall's production data show that our proposed approach has achieved better results, improving offline AUC by +1.54\%. Deployed via an efficient hybrid architecture, it achieves significant online lifts (+0.13\% UCTR, +0.25\% UCTCVR), proving its massive industrial value.
Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
arXiv:2606.04374 [cs.IR]
(or arXiv:2606.04374v1 [cs.IR] for this version)
https://doi.org/10.48550/arXiv.2606.04374 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 300 words
Related coverage
Continuous Semantic Caching for Low-Cost LLM Serving
arXiv · Oct 8, 2026
Presentation: From Models to Agents: Building Context-Aware Consumer AI at Scale at DoorDash
InfoQ · Aug 15, 2026
PaletteID: Prototype-Composed Semantic Identifiers for Multimodal CTR Prediction
arXiv · Aug 3, 2026
Asymmetric Generative Recommendation via Kronecker Residual Bridge and Multi-Faceted Hierarchical Quantization
arXiv · Aug 5, 2026