Preference Shapes Relevance: Cross-component Hierarchical Semantic Alignment for Personalized Generative Retrieval
2026-08-31 • Information Retrieval
Information RetrievalArtificial Intelligence
AI summaryⓘ
The authors present CHAP, a new method for improving how search systems match user questions with items using generative retrieval. Unlike previous methods that rely on static item descriptions, CHAP aligns the meaning of queries with item representations at multiple levels and includes user behavior to personalize results. They also speed up the search by making the generation process more efficient. Tests on multiple datasets and real-world use show that CHAP performs better than existing approaches. The authors have made their code available for others to use and build upon.
Generative RetrievalSemantic IDs (SIDs)Hierarchical Semantic AlignmentUser Behavior ModelingTransformer DecoderBeam SearchQuantizationPersonalized SearchInference LatencyMulti-granular Semantics
Authors
Gaoming Zhang, Angqing Jiang, Jianchun Song, Kena Qi, Dayao Chen, Wei Lin, Defu Lian
Abstract
Generative Retrieval (GR) has emerged as a promising paradigm by mapping queries directly to Semantic IDs (SIDs) with powerful representation capabilities for candidate items. However, existing SIDs derived solely from item content create a semantic gap, failing to align dynamic query intents with static item representations. Furthermore, current generative paradigms rarely model user behavior sequences and are always bottlenecked by the high inference latency of beam-search autoregressive decoding. To address these challenges, we propose $\textbf{C}$ross-component $\textbf{H}$ierarchical semantic $\textbf{A}$lignment for $\textbf{P}$ersonalized generative retrieval ($\textbf{CHAP}$), a novel personalized GR framework from a hierarchical perspective. First, we design a Hierarchical Semantic Alignment module to align query's latent space with item's quantization path and synchronize multi-granular semantics. Second, we construct a personalized GR framework that models user behavior by synergizing discrete SIDs for structural guidance and continuous representations for fine-grained semantic refinement. Notably, we introduce a Residual Cascading Generation mechanism to restrict the costly multi-step Transformer Decoder to a single-pass inference, boosting inference throughput while mitigating information loss. Extensive experiments on three public datasets, one proprietary industrial dataset, and online A/B tests demonstrate CHAP's superiority, validating the effectiveness and practical value of our approach. The code is publicly available at https://github.com/zzzgm/CHAP.