EDGE-Rec: Efficient and Data-Guided Edge Diffusion For Recommender Systems Graphs
This addresses the problem of improving recommendation accuracy for users by better leveraging available data, though it appears incremental as it builds on existing collaborative filtering and diffusion principles.
The paper tackles the under-utilization of user features, item features, and interaction strengths in recommender systems by proposing a Row-Column Separable Attention mechanism and Graph Diffusion Transformer architecture that directly denoises weighted interaction matrices, achieving user-item rating predictions on the original rating scale.
Most recommender systems research focuses on binary historical user-item interaction encodings to predict future interactions. User features, item features, and interaction strengths remain largely under-utilized in this space or only indirectly utilized, despite proving largely effective in large-scale production recommendation systems. We propose a new attention mechanism, loosely based on the principles of collaborative filtering, called Row-Column Separable Attention RCSA to take advantage of real-valued interaction weights as well as user and item features directly. Building on this mechanism, we additionally propose a novel Graph Diffusion Transformer GDiT architecture which is trained to iteratively denoise the weighted interaction matrix of the user-item interaction graph directly. The weighted interaction matrix is built from the bipartite structure of the user-item interaction graph and corresponding edge weights derived from user-item rating interactions. Inspired by the recent progress in text-conditioned image generation, our method directly produces user-item rating predictions on the same scale as the original ratings by conditioning the denoising process on user and item features with a principled approach.