Towards A Tri-View Diffusion Framework for Recommendation

TL;DR

This paper introduces a novel diffusion framework for recommendation that balances energy and entropy considerations, employing a thermodynamic perspective, a specialized denoiser, and an adaptive sampling process, leading to improved accuracy and efficiency.

Contribution

It proposes a minimalistic diffusion model incorporating thermodynamic principles, a tailored denoiser for anisotropy preservation, and an acceptance-rejection Gumbel sampling process for enhanced recommendation performance.

Findings

Outperforms baselines in accuracy and efficiency.

Reveals energy maximization in DMs contrasts with entropy reduction in classic models.

Demonstrates the effectiveness of the proposed thermodynamic diffusion framework.

Abstract

Diffusion models (DMs) have recently gained significant interest for their exceptional potential in recommendation tasks. This stems primarily from their prominent capability in distilling, modeling, and generating comprehensive user preferences. However, previous work fails to examine DMs in recommendation tasks through a rigorous lens. In this paper, we first experimentally investigate the completeness of recommender models from a thermodynamic view. We reveal that existing DM-based recommender models operate by maximizing the energy, while classic recommender models operate by reducing the entropy. Based on this finding, we propose a minimalistic diffusion framework that incorporates both factors via the maximization of Helmholtz free energy. Meanwhile, to foster the optimization, our reverse process is armed with a well-designed denoiser to maintain the inherent anisotropy, which measures the user-item cross-correlation in the context of bipartite graphs. Finally, we adopt an Acceptance-Rejection Gumbel Sampling Process (AR-GSP) to prioritize the far-outnumbered unobserved interactions for model robustness. AR-GSP integrates an acceptance-rejection sampling to ensure high-quality hard negative samples for general recommendation tasks, and a timestep-dependent Gumbel Softmax to handle an adaptive sampling strategy for diffusion models. Theoretical analyses and extensive experiments demonstrate that our proposed framework has distinct superiority over baselines in terms of accuracy and efficiency.