Behavior-Dependent Linear Recurrent Units for Efficient Sequential Recommendation

TL;DR

RecBLR introduces a behavior-dependent linear recurrent unit model that achieves efficient training, low-cost inference, and high recommendation performance simultaneously, addressing key challenges in sequential recommendation systems.

Contribution

The paper proposes RecBLR, a novel linear recurrent unit model with behavior-dependent mechanisms and hardware-aware acceleration, achieving the three golden principles of sequential recommendation.

Findings

RecBLR outperforms existing models in recommendation accuracy.

RecBLR demonstrates significant training and inference efficiency.

The model scales effectively to long user interaction sequences.

Abstract

Sequential recommender systems aims to predict the users' next interaction through user behavior modeling with various operators like RNNs and attentions. However, existing models generally fail to achieve the three golden principles for sequential recommendation simultaneously, i.e., training efficiency, low-cost inference, and strong performance. To this end, we propose RecBLR, an Efficient Sequential Recommendation Model based on Behavior-Dependent Linear Recurrent Units to accomplish the impossible triangle of the three principles. By incorporating gating mechanisms and behavior-dependent designs into linear recurrent units, our model significantly enhances user behavior modeling and recommendation performance. Furthermore, we unlock the parallelizable training as well as inference efficiency for our model by designing a hardware-aware scanning acceleration algorithm with a customized CUDA kernel. Extensive experiments on real-world datasets with varying lengths of user behavior sequences demonstrate RecBLR's remarkable effectiveness in simultaneously achieving all three golden principles - strong recommendation performance, training efficiency, and low-cost inference, while exhibiting excellent scalability to datasets with long user interaction histories.