Journal of Computer Applications ›› 2024, Vol. 44 ›› Issue (9): 2719-2725.DOI: 10.11772/j.issn.1001-9081.2023091255

• Data science and technology • Previous Articles     Next Articles

Multi-layer information interactive fusion algorithm based on graph neural network for session-based recommendation

Hang YANG, Wanggen LI(), Gensheng ZHANG, Zhige WANG, Xin KAI   

  1. School of Computer and Information,Anhui Normal University,Wuhu Anhui 241003,China
  • Received:2023-09-18 Revised:2023-12-20 Accepted:2023-12-22 Online:2024-02-20 Published:2024-09-10
  • Contact: Wanggen LI
  • About author:YANG Hang, born in 1999, M. S. candidate. His research interests include deep learning, recommender system.
    ZHANG Gensheng, born in 1999, M. S. candidate. His research interests include recommender system, sequential recommendation, deep learning.
    WANG Zhige, born in 1997, M. S. candidate. His research interests include recommender system, advertising calculation, deep learning.
    KAI Xin, born in 2000, M. S. candidate. His research interests include recommender system, deep learning.
  • Supported by:
    National Natural Science Foundation of China(61976006)

基于图神经网络的多层信息交互融合算法用于会话推荐

杨航, 李汪根(), 张根生, 王志格, 开新   

  1. 安徽师范大学 计算机与信息学院,安徽 芜湖 241003
  • 通讯作者: 李汪根
  • 作者简介:杨航(1999—),男,安徽池州人,硕士研究生,主要研究方向:深度学习、推荐系统
    李汪根(1973—),男,安徽太湖人,教授,博士,主要研究方向:生物计算、智能计算
    张根生(1999—),男,安徽桐城人,硕士研究生,主要研究方向:推荐系统、序列推荐、深度学习
    王志格(1997—),男,安徽池州人,硕士研究生,主要研究方向:推荐系统、广告计算、深度学习
    开新(2000—),男,安徽合肥人,硕士研究生,主要研究方向:推荐系统、深度学习。
  • 基金资助:
    国家自然科学基金资助项目(61976006)

Abstract:

Addressing the insufficient exploration of item-transition information within the current session and the limited utilization of other session details in session-based recommendation nowadays, a multi-layer information interactive fusion algorithm based on graph neural network was proposed for session-based recommendation. Based on the current session, firstly, the information of neighborhood nodes was aggregated by designing different weights for the connection relationships between nodes, and the explicit information of item-transition in the current session was mined. Secondly, the neighborhood node information was aggregated by stacked residual graph attention network, and the implicit item-transition information in the current session was mined. Finally, the sequence-dependent information in the time stamp-based session was mined through a single gated graph neural network. Based on other sessions, the entire set of sessions was linked through the first-order neighbors of nodes, and the global information encoding was learnt, and then, the embedding representations of four levels were integrated to obtain more comprehensive item-transition information. At the same time, soft attention mechanism and reverse position embedding information were used to fuse the obtained item-transition information more effectively. Experimental results show that the precision P@20 and mean reciprocal rank MRR@20 of the proposed algorithm are increased by 0.79% and 0.84% respectively compared with the suboptimal model GCE-GNN (Global Context Enhanced Graph Neural Network) on Diginetica dataset, the P@20 and MRR@20 of the proposed algorithm are increased by 8.23% and 7.86% respectively compared with the suboptimal model HyperS2Rec on Tmall dataset, and the P@20 and MRR@20 of the proposed algorithm are increased by 1.33% and 7.16% respectively compared with the suboptimal model HyperS2Rec on Nowplaying dataset.

Key words: session-based recommendation, residual graph attention network, gated graph neural network, soft attention, reverse position embedding

摘要:

针对当前会话推荐中存在对于当前会话的项目转换信息挖掘不充分且极少利用其他会话信息的问题,提出一种基于图神经网络的多层信息交互融合算法用于会话推荐。基于当前会话,首先,对节点之间的连接关系设计不同的权重聚合邻域节点的信息,并挖掘当前会话中项目转换的显性信息;其次,通过基于堆叠的残差图注意力网络聚合邻域节点信息,挖掘当前会话中项目转换的隐性信息;最后,通过单门控图神经网络挖掘基于时间戳的会话中存在的序列依赖信息。基于其他会话,通过节点的一阶邻居将整个会话集联系起来,学习全局信息编码,进而融合4个层次的嵌入表示以获得更全面的项目转换信息,同时使用软注意力机制和反向位置嵌入信息对获得的项目转换信息进行更有效的融合。实验结果表明,在Diginetica数据集上,所提模型的精度P@20和平均倒数排名MRR@20较次优模型GCE-GNN(Global Context Enhanced Graph Neural Network)分别提升了0.79%和0.84%;在Tmall数据集上,所提模型的P@20和MRR@20较次优模型HyperS2Rec分别提升了8.23%和7.86%;在Nowplaying数据集上,所提模型的P@20和MRR@20较次优模型HyperS2Rec分别提升了1.33%和7.16%。

关键词: 会话推荐, 残差图注意力网络, 门控图神经网络, 软注意力, 反向位置嵌入

CLC Number: