Dual-tower model with semantic perception and timespan-coupled hypergraph for next-basket recommendation.

Autor: Zhou Y; School of Computer Science and Technology, Xidian University, Xi'an, 710071, China; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China. Electronic address: zhou_yt@stu.xidian.edu.cn., Chu H; School of Computer Science and Technology, Xidian University, Xi'an, 710071, China; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China. Electronic address: hchu@mail.xidian.edu.cn., Li Q; School of Computer Science and Technology, Xidian University, Xi'an, 710071, China; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China. Electronic address: qshli@mail.xidian.edu.cn., Li J; School of Computer Science and Technology, Xidian University, Xi'an, 710071, China; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China. Electronic address: lijianan@xidian.edu.cn., Zhang S; School of Computer Science and Technology, Xidian University, Xi'an, 710071, China; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China. Electronic address: zhang_s@stu.xidian.edu.cn., Zhu F; School of Computer Science and Technology, Xidian University, Xi'an, 710071, China; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China. Electronic address: ff.zhu@stu.xidian.edu.cn., Hu J; School of Computer Science and Technology, Xidian University, Xi'an, 710071, China; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China. Electronic address: hujingzhao@xidian.edu.cn., Wang L; School of Computer Science and Technology, Xidian University, Xi'an, 710071, China; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China. Electronic address: wangluqiao@stu.xidian.edu.cn., Yang W; Intelligent Financial Software Engineering New Technology Joint Laboratory, Xidian University, Xi'an, 710071, China; Shanghai Fairyland Software Corp., Ltd., Shanghai, 200233, China. Electronic address: zhaoxin@fulan.com.cn.
Jazyk: angličtina
Zdroj: Neural networks : the official journal of the International Neural Network Society [Neural Netw] 2024 Dec 05; Vol. 184, pp. 107001. Date of Electronic Publication: 2024 Dec 05.
DOI: 10.1016/j.neunet.2024.107001
Abstrakt: Next basket recommendation (NBR) is an essential task within the realm of recommendation systems and is dedicated to the anticipation of user preferences in the next moment based on the analysis of users' historical sequences of engaged baskets. Current NBR models utilise unique identity (ID) information to represent distinct users and items and focus on capturing the dynamic preferences of users through sequential encoding techniques such as recurrent neural networks and hierarchical time decay modelling, which have dominated the NBR field more than a decade. However, these models exhibit two significant limitations, resulting in suboptimal representations for both users and items. First, the dependence on unique ID information for the derivation of user and item representations ignores the rich semantic relations that interweave the items. Second, the majority of NBR models remain bound to model an individual user's historical basket sequence, thereby neglecting the broader vista of global collaborative relations among users and items. To address these limitations, we introduce a dual-tower model with semantic perception and timespan-coupled hypergraph for the NBR. It is carefully designed to integrate semantic and collaborative relations into both user and item representations. Specifically, to capture rich semantic relations effectively, we propose a hierarchical semantic attention mechanism with a large language model to integrate multi-aspect textual semantic features of items for basket representation learning. Simultaneously, to capture global collaborative relations explicitly, we design a timespan-coupled hypergraph convolutional network to efficiently model high-order structural connectivity on a hypergraph among users and items. Finally, a multi-objective joint optimisation loss is used to optimise the learning and integration of semantic and collaborative relations for recommendation. Comprehensive experiments on two public datasets demonstrate that our proposed model significantly outperforms the mainstream NBR models on two classical evaluation metrics, Recall and Normalised Discounted Cumulative Gain (NDCG).
Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(Copyright © 2024. Published by Elsevier Ltd.)
Databáze: MEDLINE
Externí odkaz: