A Systematic Reproducibility Study of BSARec for Sequential Recommendation

TL;DR

This paper systematically reproduces and evaluates BSARec, a Transformer-based sequential recommendation model that incorporates frequency domain techniques, revealing insights into its effectiveness and the impact of various implementation choices.

Contribution

It provides a thorough validation of BSARec's performance, compares different frequency domain methods, and analyzes padding strategies to improve high-frequency signal capture in SR models.

Findings

BSARec outperforms some baseline SR methods on certain datasets.

Discrete wavelet transform offers slight improvements over Fourier transform.

Non-constant padding significantly enhances recommendation accuracy.

Abstract

In sequential recommendation (SR), the self-attention mechanism of Transformer-based models acts as a low-pass filter, limiting their ability to capture high-frequency signals that reflect short-term user interests. To overcome this, BSARec augments the Transformer encoder with a frequency layer that rescales high-frequency components using the Fourier transform. However, the overall effectiveness of BSARec and the roles of its individual components have yet to be systematically validated. We reproduce BSARec and show that it outperforms other SR methods on some datasets. To empirically assess whether BSARec improves performance on high-frequency signals, we propose a metric to quantify user history frequency and evaluate SR methods across different user groups. We compare digital signal processing (DSP) techniques and find that the discrete wavelet transform (DWT) offer only slight improvements over Fourier transforms, and DSP methods provide no clear advantage over simple residual connections. Finally, we explore padding strategies and find that non-constant padding significantly improves recommendation performance, whereas constant padding hinders the frequency rescaler's ability to capture high-frequency signals.