Multi-Hop Multi-View Memory Transformer for Session-Based Recommendation

Xingrui Zhuo

doi:10.1145/3663760

Multi-Hop Multi-View Memory Transformer for Session-Based Recommendation

2024-07-11

0

OA

AI

X

Xingrui Zhuo

钱

钱胜胜 (Shengsheng Qian)

DOI：10.1145/3663760

Original request for help

Save

Abstract

En 中文

A Session-Based Recommendation (SBR) seeks to predict users' future item preferences by analyzing their interactions with previously clicked items. In recent approaches, Graph Neural Networks (GNNs) have been commonly applied to capture item relations within a session to infer user intentions. However, these GNNbased methods typically struggle with feature ambiguity between the sequential session information and the item conversion within an item graph, which may impede the model's ability to accurately infer user intentions. In this article, we propose a novel Multi-hop Multi-view Memory Transformer (M3T) to effectively integrate the sequence-view information and relation conversion (graph-view information) of items in a session. First, we propose a Multi-view Memory Transformer (M2T) module to concurrently obtain multi-view information of items. Then, a set of trainable memory matrices are employed to store sharable item features, which mitigates cross-view item feature ambiguity. To comprehensively capture latent user intentions, an M3T framework is designed to integrate user intentions across different hops of an item graph. Specifically, a k-order power method is proposed to manage the item graph to alleviate the over-smoothing problem when obtaining high-order relations of items. Extensive experiments conducted on three real-world datasets demonstrate the superiority of our method.

Keywords:

Session-based recommendation

multi-view intention fusion

memory Transformer

multi-hop graph embedding

Multi-Hop Multi-View Memory Transformer for Session-Based Recommendation

Abstract

Journal

Organization