A Simple and Effective Framework for Symmetric Consistent Indexing in Large-Scale Dense Retrieval

2025-12-15

AI Summary

To address representation space misalignment and training-inference inconsistency in dual-tower dense retrieval—which degrade long-tail query performance, impair semantic ID quality, and hinder generative recommendation—this paper proposes the SCI framework. First, an input-swapping mechanism enables parameter-free, symmetric representation space alignment. Second, a dual-perspective indexing strategy ensures strict training-inference consistency, with theoretical guarantees, zero additional parameters, and engineering efficiency. Third, coarse-grained and fine-grained semantic ID modeling is jointly incorporated. Evaluated on public and e-commerce billion-scale datasets, SCI significantly improves recall and retrieval stability—especially for long-tail queries—while enabling millisecond-level billion-scale deployment.

Technology Category

• Search and Optimization: Learning to Search
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs

Abstract

Dense retrieval has become the industry standard in large-scale information retrieval systems due to its high efficiency and competitive accuracy. Its core relies on a coarse-to-fine hierarchical architecture that enables rapid candidate selection and precise semantic matching, achieving millisecond-level response over billion-scale corpora. This capability makes it essential not only in traditional search and recommendation scenarios but also in the emerging paradigm of generative recommendation driven by large language models, where semantic IDs-themselves a form of coarse-to-fine representation-play a foundational role. However, the widely adopted dual-tower encoding architecture introduces inherent challenges, primarily representational space misalignment and retrieval index inconsistency, which degrade matching accuracy, retrieval stability, and performance on long-tail queries. These issues are further magnified in semantic ID generation, ultimately limiting the performance ceiling of downstream generative models. To address these challenges, this paper proposes a simple and effective framework named SCI comprising two synergistic modules: a symmetric representation alignment module that employs an innovative input-swapping mechanism to unify the dual-tower representation space without adding parameters, and an consistent indexing with dual-tower synergy module that redesigns retrieval paths using a dual-view indexing strategy to maintain consistency from training to inference. The framework is systematic, lightweight, and engineering-friendly, requiring minimal overhead while fully supporting billion-scale deployment. We provide theoretical guarantees for our approach, with its effectiveness validated by results across public datasets and real-world e-commerce datasets.

Problem

• Addresses representational space misalignment in dual-tower dense retrieval
• Resolves retrieval index inconsistency from training to inference
• Improves accuracy and stability for long-tail queries in large-scale systems

Innovation

• Symmetric representation alignment with input-swapping mechanism
• Dual-view indexing strategy for training-inference consistency
• Lightweight framework for billion-scale deployment without added parameters