Exploiting Session Information in BERT-based Session-aware Sequential Recommendation

TL;DR

This paper enhances BERT-based sequential recommendation models by incorporating session information through simple, parameter-efficient methods like session tokens, segment embeddings, and time-aware attention, leading to improved performance.

Contribution

The paper introduces three novel, lightweight techniques to exploit session data in BERT-based recommendation models, reducing complexity while boosting accuracy.

Findings

Improved recommendation accuracy on benchmark datasets.

Effective integration of session information with minimal additional parameters.

Validation of proposed methods through extensive experiments.

Abstract

In recommendation systems, utilizing the user interaction history as sequential information has resulted in great performance improvement. However, in many online services, user interactions are commonly grouped by sessions that presumably share preferences, which requires a different approach from ordinary sequence representation techniques. To this end, sequence representation models with a hierarchical structure or various viewpoints have been developed but with a rather complex network structure. In this paper, we propose three methods to improve recommendation performance by exploiting session information while minimizing additional parameters in a BERT-based sequential recommendation model: using session tokens, adding session segment embeddings, and a time-aware self-attention. We demonstrate the feasibility of the proposed methods through experiments on widely used recommendation datasets.