Adaptive Item-based Collaborative Structures via Noise Rescheduling in Diffusion for Generative Recommendation
2026-08-24 • Information Retrieval
Information RetrievalArtificial Intelligence
AI summaryⓘ
The authors study recommendation systems that use Discrete Diffusion Models (DDMs), which generate user history step-by-step but often miss important item relationships. To fix this, they create ANR-DiffRec, a new method that adds information about how items are related (using an item co-occurrence matrix) and adjusts the noise in the model based on item connections. This helps the model better understand item patterns during training. Their tests show this approach improves recommendation quality compared to existing methods.
Discrete Diffusion ModelsRecommendation SystemsItem Co-occurrence MatrixCollaborative FilteringNoise ScheduleDenoising ProcessGenerative ModelsSequential PatternsAdaptive Noise Rescheduling
Authors
Jiaqi Wang, Tianying Liu, Heng Chang, Jihong Guan, Wengen Li, Shuigeng Zhou
Abstract
Discrete Diffusion Models (DDMs) have recently been introduced to recommendation systems, modeling user history as a token generation process via iterative denoising. However, while effective at capturing user-level sequential patterns, these methods often fail to explicitly integrate item-based collaborative filtering information, a critical component for accurate recommendation. This deficiency manifests in two key aspects: (1) the item representation is often semantic-focused, lacking collaborative priors for diffusion training; and (2) the denoising process employs a uniform noise schedule, treating all tokens indiscriminately and ignoring item-level adaptive structural dependencies. To bridge this gap, we propose ANR-DiffRec, a unified framework designed to encode item-based collaborative structures into discrete diffusion for generative recommendation. First, we explicitly incorporate an item co-occurrence matrix to guide semantic ID generation, providing a structured collaborative prior for discrete diffusion training. Second, we introduce an item-based adaptive noise rescheduling mechanism that dynamically adjusts denoising weights according to both local contextual recoverability and behavior-aware item dependencies. Specifically, the proposed strategy jointly models intra-item structural context and inter-item collaborative signals, enabling structure-aware denoising during diffusion training. Extensive experiments on multiple benchmarks demonstrate that our method consistently outperforms state-of-the-art generative recommendation models. Code: https://github.com/CalmaQi/ANR-DiffRec.