Incorporating Classifier-Free Guidance in Diffusion Model-Based Recommendation
Noah Buchanan, Susan Gauch, Quan Mai
TL;DR
The paper addresses improving diffusion-based recommender systems by integrating classifier-free guidance to reduce reliance on external classifiers and enhance performance, especially when data are sparse. It introduces a training setup that jointly learns unconditional and conditional denoisers within a single network and uses pre-noise guidance with a fixed unconditional probability to guide recommendations. The approach is evaluated on MovieLens 1M, Yelp, and Amazon datasets, showing that the classifier-free guided diffusion recommender generally matches or surpasses prior diffusion-based methods across multiple metrics, with strong gains in sparse settings and few-shot scenarios. The work highlights practical implications for diffusion-based recommendations and outlines future directions, including transformer backbones and refined conditioning, to further boost performance and applicability in real-world systems.
Abstract
This paper presents a diffusion-based recommender system that incorporates classifier-free guidance. Most current recommender systems provide recommendations using conventional methods such as collaborative or content-based filtering. Diffusion is a new approach to generative AI that improves on previous generative AI approaches such as Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs). We incorporate diffusion in a recommender system that mirrors the sequence users take when browsing and rating items. Although a few current recommender systems incorporate diffusion, they do not incorporate classifier-free guidance, a new innovation in diffusion models as a whole. In this paper, we present a diffusion recommender system that augments the underlying recommender system model for improved performance and also incorporates classifier-free guidance. Our findings show improvements over state-of-the-art recommender systems for most metrics for several recommendation tasks on a variety of datasets. In particular, our approach demonstrates the potential to provide better recommendations when data is sparse.