Masked Diffusion Generative Recommendation

TL;DR

This paper introduces MDGR, a diffusion-based recommendation framework that improves efficiency and accuracy over traditional autoregressive methods by utilizing a parallel decoding strategy and adaptive masking, leading to better performance and practical benefits.

Contribution

The paper proposes a novel diffusion-based recommendation model with a parallel codebook, adaptive masking during training, and a two-stage decoding process, addressing limitations of autoregressive approaches.

Findings

MDGR outperforms 10 state-of-the-art baselines by up to 10.78%.

Deployment on a large-scale platform increased revenue by 1.20%.

The method achieves better efficiency and recommendation quality.

Abstract

Generative recommendation (GR) typically first quantizes continuous item embeddings into multi-level semantic IDs (SIDs), and then generates the next item via autoregressive decoding. Although existing methods are already competitive in terms of recommendation performance, directly inheriting the autoregressive decoding paradigm from language models still suffers from three key limitations: (1) autoregressive decoding struggles to jointly capture global dependencies among the multi-dimensional features associated with different positions of SID; (2) using a unified, fixed decoding path for the same item implicitly assumes that all users attend to item attributes in the same order; (3) autoregressive decoding is inefficient at inference time and struggles to meet real-time requirements. To tackle these challenges, we propose MDGR, a Masked Diffusion Generative Recommendation framework that reshapes the GR pipeline from three perspectives: codebook, training, and inference. (1) We adopt a parallel codebook to provide a structural foundation for diffusion-based GR. (2) During training, we adaptively construct masking supervision signals along both the temporal and sample dimensions. (3) During inference, we develop a warm-up-based two-stage parallel decoding strategy for efficient generation of SIDs. Extensive experiments on multiple public and industrial-scale datasets show that MDGR outperforms ten state-of-the-art baselines by up to 10.78%. Furthermore, by deploying MDGR on a large-scale online advertising platform, we achieve a 1.20% increase in revenue, demonstrating its practical value.