An Attentive Inductive Bias for Sequential Recommendation beyond the Self-Attention

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Shin, Yehjin, Choi, Jeongwhan, Wi, Hyowon, Park, Noseong
Format: Preprint
Publié: 2023
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917591452221440
author Shin, Yehjin
Choi, Jeongwhan
Wi, Hyowon
Park, Noseong
author_facet Shin, Yehjin
Choi, Jeongwhan
Wi, Hyowon
Park, Noseong
contents Sequential recommendation (SR) models based on Transformers have achieved remarkable successes. The self-attention mechanism of Transformers for computer vision and natural language processing suffers from the oversmoothing problem, i.e., hidden representations becoming similar to tokens. In the SR domain, we, for the first time, show that the same problem occurs. We present pioneering investigations that reveal the low-pass filtering nature of self-attention in the SR, which causes oversmoothing. To this end, we propose a novel method called $\textbf{B}$eyond $\textbf{S}$elf-$\textbf{A}$ttention for Sequential $\textbf{Rec}$ommendation (BSARec), which leverages the Fourier transform to i) inject an inductive bias by considering fine-grained sequential patterns and ii) integrate low and high-frequency information to mitigate oversmoothing. Our discovery shows significant advancements in the SR domain and is expected to bridge the gap for existing Transformer-based SR models. We test our proposed approach through extensive experiments on 6 benchmark datasets. The experimental results demonstrate that our model outperforms 7 baseline methods in terms of recommendation performance. Our code is available at https://github.com/yehjin-shin/BSARec.
format Preprint
id arxiv_https___arxiv_org_abs_2312_10325
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle An Attentive Inductive Bias for Sequential Recommendation beyond the Self-Attention
Shin, Yehjin
Choi, Jeongwhan
Wi, Hyowon
Park, Noseong
Machine Learning
Artificial Intelligence
Information Retrieval
Sequential recommendation (SR) models based on Transformers have achieved remarkable successes. The self-attention mechanism of Transformers for computer vision and natural language processing suffers from the oversmoothing problem, i.e., hidden representations becoming similar to tokens. In the SR domain, we, for the first time, show that the same problem occurs. We present pioneering investigations that reveal the low-pass filtering nature of self-attention in the SR, which causes oversmoothing. To this end, we propose a novel method called $\textbf{B}$eyond $\textbf{S}$elf-$\textbf{A}$ttention for Sequential $\textbf{Rec}$ommendation (BSARec), which leverages the Fourier transform to i) inject an inductive bias by considering fine-grained sequential patterns and ii) integrate low and high-frequency information to mitigate oversmoothing. Our discovery shows significant advancements in the SR domain and is expected to bridge the gap for existing Transformer-based SR models. We test our proposed approach through extensive experiments on 6 benchmark datasets. The experimental results demonstrate that our model outperforms 7 baseline methods in terms of recommendation performance. Our code is available at https://github.com/yehjin-shin/BSARec.
title An Attentive Inductive Bias for Sequential Recommendation beyond the Self-Attention
topic Machine Learning
Artificial Intelligence
Information Retrieval
url https://arxiv.org/abs/2312.10325