Sequential Recommendation with Relation-Aware Kernelized Self-Attention

TL;DR

This paper introduces Relation-Aware Kernelized Self-Attention (RKSA), a novel probabilistic self-attention model for sequential recommendation that enhances interpretability and performance by modeling relations in a latent space.

Contribution

RKSA combines Transformer self-attention with a probabilistic latent space to incorporate relation-awareness in sequential recommendation tasks.

Findings

RKSA outperforms recent baseline models on benchmark datasets.

RKSA provides interpretable latent space explanations for recommendations.

RKSA significantly improves recommendation accuracy.

Abstract

Recent studies identified that sequential Recommendation is improved by the attention mechanism. By following this development, we propose Relation-Aware Kernelized Self-Attention (RKSA) adopting a self-attention mechanism of the Transformer with augmentation of a probabilistic model. The original self-attention of Transformer is a deterministic measure without relation-awareness. Therefore, we introduce a latent space to the self-attention, and the latent space models the recommendation context from relation as a multivariate skew-normal distribution with a kernelized covariance matrix from co-occurrences, item characteristics, and user information. This work merges the self-attention of the Transformer and the sequential recommendation by adding a probabilistic model of the recommendation task specifics. We experimented RKSA over the benchmark datasets, and RKSA shows significant improvements compared to the recent baseline models. Also, RKSA were able to produce a latent space model that answers the reasons for recommendation.