HiD-VAE: Interpretable Generative Recommendation via Hierarchical and Disentangled Semantic IDs

2025-08-06

AI Summary

Existing generative recommendation methods suffer from flat, uninterpretable semantic IDs and frequent ID collisions due to unsupervised tokenization, degrading both accuracy and diversity. To address this, we propose HiD-VAE—a variational autoencoder-based framework introducing the first hierarchical supervised quantization mechanism. It constructs a discrete codebook capable of predicting multi-level semantic labels and incorporates a uniqueness loss to mitigate latent space overlap, enabling hierarchical modeling and disentangled representation of semantic IDs. This design ensures traceable and interpretable recommendation generation. Evaluated on three public benchmarks, HiD-VAE consistently outperforms state-of-the-art methods, achieving simultaneous improvements in recommendation accuracy and diversity while significantly enhancing model interpretability.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Generation
• Data Mining & Knowledge Management: Recommender Systems

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Recommender systems are indispensable for helping users navigate the immense item catalogs of modern online platforms. Recently, generative recommendation has emerged as a promising paradigm, unifying the conventional retrieve-and-rank pipeline into an end-to-end model capable of dynamic generation. However, existing generative methods are fundamentally constrained by their unsupervised tokenization, which generates semantic IDs suffering from two critical flaws: (1) they are semantically flat and uninterpretable, lacking a coherent hierarchy, and (2) they are prone to representation entanglement (i.e., ``ID collisions''), which harms recommendation accuracy and diversity. To overcome these limitations, we propose HiD-VAE, a novel framework that learns hierarchically disentangled item representations through two core innovations. First, HiD-VAE pioneers a hierarchically-supervised quantization process that aligns discrete codes with multi-level item tags, yielding more uniform and disentangled IDs. Crucially, the trained codebooks can predict hierarchical tags, providing a traceable and interpretable semantic path for each recommendation. Second, to combat representation entanglement, HiD-VAE incorporates a novel uniqueness loss that directly penalizes latent space overlap. This mechanism not only resolves the critical ID collision problem but also promotes recommendation diversity by ensuring a more comprehensive utilization of the item representation space. These high-quality, disentangled IDs provide a powerful foundation for downstream generative models. Extensive experiments on three public benchmarks validate HiD-VAE's superior performance against state-of-the-art methods. The code is available at https://anonymous.4open.science/r/HiD-VAE-84B2.

Problem

• Generative recommendation lacks interpretable hierarchical semantic IDs
• Existing methods suffer from representation entanglement and ID collisions
• Proposing HiD-VAE for disentangled and interpretable item representations

Innovation

• Hierarchically-supervised quantization for interpretable IDs
• Uniqueness loss to prevent representation entanglement
• Disentangled item representations for better recommendations