Adaptive Diffusion-based Augmentation for Recommendation

2026-01-04

arXiv.org

AI Summary

This work addresses the challenges in recommendation systems where negative sampling often mislabels potential positive items and lacks fine-grained control over negative sample difficulty. To this end, the authors propose a model-agnostic diffusion-based negative sample generation module. This approach introduces diffusion models to recommender systems for the first time, simulating a continuous transition process from positive to negative samples. By integrating a theoretically grounded transition point identification mechanism with a score-aware function, the method adaptively selects the optimal diffusion timestep to enable precise control over negative sample hardness. Extensive experiments demonstrate that the proposed module significantly enhances performance across mainstream architectures—including collaborative filtering and sequential recommendation—without requiring any modifications to the original models, thereby exhibiting strong compatibility and effectiveness.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Data Mining & Knowledge Management: Recommender Systems
• Computer Vision: Diffusion Models for Vision

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Content-based information diffusion

Abstract

Recommendation systems often rely on implicit feedback, where only positive user-item interactions can be observed. Negative sampling is therefore crucial to provide proper negative training signals. However, existing methods tend to mislabel potentially positive but unobserved items as negatives and lack precise control over negative sample selection. We aim to address these by generating controllable negative samples, rather than sampling from the existing item pool. In this context, we propose Adaptive Diffusion-based Augmentation for Recommendation (ADAR), a novel and model-agnostic module that leverages diffusion to synthesize informative negatives. Inspired by the progressive corruption process in diffusion, ADAR simulates a continuous transition from positive to negative, allowing for fine-grained control over sample hardness. To mine suitable negative samples, we theoretically identify the transition point at which a positive sample turns negative and derive a score-aware function to adaptively determine the optimal sampling timestep. By identifying this transition point, ADAR generates challenging negative samples that effectively refine the model's decision boundary. Experiments confirm that ADAR is broadly compatible and boosts the performance of existing recommendation models substantially, including collaborative filtering and sequential recommendation, without architectural modifications.

Problem

• negative sampling
• implicit feedback
• recommendation systems
• sample hardness
• diffusion-based augmentation

Innovation

• diffusion-based augmentation
• negative sampling
• recommendation systems
• adaptive sampling
• implicit feedback