RecFusion: A Binomial Diffusion Process for 1D Data for Recommendation

TL;DR

RecFusion introduces a novel binomial diffusion process tailored for 1D binary user-item interaction data, achieving competitive recommendation performance and offering potential applications in medical imaging.

Contribution

It proposes a new binomial diffusion model specifically designed for 1D binary data in recommendation systems, filling a gap in existing diffusion approaches.

Findings

RecFusion approaches VAE baseline performance on recommendation datasets.

The model effectively handles binary, non-sequential user-item interactions.

Potential applications extend to medical imaging modalities.

Abstract

In this paper we propose RecFusion, which comprise a set of diffusion models for recommendation. Unlike image data which contain spatial correlations, a user-item interaction matrix, commonly utilized in recommendation, lacks spatial relationships between users and items. We formulate diffusion on a 1D vector and propose binomial diffusion, which explicitly models binary user-item interactions with a Bernoulli process. We show that RecFusion approaches the performance of complex VAE baselines on the core recommendation setting (top-n recommendation for binary non-sequential feedback) and the most common datasets (MovieLens and Netflix). Our proposed diffusion models that are specialized for 1D and/or binary setups have implications beyond recommendation systems, such as in the medical domain with MRI and CT scans.