面向捆绑推荐的解耦图对比学习

doi:10.3778/j.issn.1002-8331.2403-0149

摘要/Abstract

摘要： 捆绑推荐的目的是将多个相关的项目作为一个整体推荐给用户。在捆绑推荐中，用户可能会因为捆绑包中的一个特定项目去选择该捆绑包，而用户选择项目往往包含多种意图。现有的方法主要是对用户和捆绑包的意图解耦表示进行整体建模，而忽略了不同意图之间的细微差别。设计了一个面向捆绑推荐的解耦图对比学习模型（disentangled graph contrastive learning for bundle recommendation，DCBR），解耦用户的潜在意图，并通过对比学习在宏观视图和微观视图之间构建合作关联。具体来说，分别从宏观视图和微观视图学习用户意图，并生成意图解耦表示。根据捆绑包-项目隶属关系图分别映射出用户对捆绑包的偏好。设计了一种意图级对比学习在每个意图子空间中构建两个视图的合作关联。在三个公共数据集上的大量实验表明，所提出的模型要优于基线模型，该模型在Recall和NDCG指标上分别比最佳基线提高了3.26%~6.14%、3.99%~6.78%和4.79%~7.97%。

关键词: 捆绑推荐, 解耦表示学习, 对比学习, 图神经网络

Abstract: The goal of bundle recommendation is to recommend multiple related items to users as a whole. In bundle recommendation, a user may select a bundle because of a specific item in the bundle, and users choose items often with multiple intents. Existing methods mainly model the intent disentangled representations of user and bundle as a whole, while ignoring the subtle differences among different intents. To address this problem, a disentangled graph contrastive learning model for bundle recommendation (DCBR) is proposed, which disentangles the potential intents of the user and constructs a cooperative association between macro view and micro view through contrastive learning. More specifically, this paper begins with learning user intents from macro view and micro view, respectively, and generates the intent disentangled representations. Then, the preferences of user for bundles are mapped according to the bundle-item affiliation graph. Finally, intent-level contrastive learning is designed to build cooperative associations created by two views in each intent subspace. Extensive experiments on three public datasets show that the proposed model outperforms the state-of-the-art baselines, the model improves over the best baseline by 3.26% to 6.14%, 3.99% to 6.78%, and 4.79% to 7.97% on Recall and NDCG metrics, respectively.

Key words: bundle recommendation, disentangled representation learning, contrastive learning, graph neural networks

张尧, 王绍卿, 吴瑕, 孙福振. 面向捆绑推荐的解耦图对比学习[J]. 计算机工程与应用, 2025, 61(12): 154-165.

ZHANG Yao, WANG Shaoqing, WU Xia, SUN Fuzhen. Disentangled Graph Contrastive Learning for Bundle Recommendation[J]. Computer Engineering and Applications, 2025, 61(12): 154-165.

参考文献

[1] GE Y Q, ZHAO S Y, ZHOU H L, et al. Understanding echo chambers in E-commerce recommender systems[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2020: 2261-2270.
[2] KERSBERGEN B, SCHELTER S. Learnings from a retail recommendation system on billions of interactions at bol. com[C]//Proceedings of the 2021 IEEE 37th International Conference on Data Engineering. Piscataway: IEEE, 2021: 2447-2452.
[3] 李宇琦, 陈维政, 闫宏飞, 等. 基于网络表示学习的个性化商品推荐[J]. 计算机学报, 2019, 42(8): 1767-1778.
LI Y Q, CHEN W Z, YAN H F, et al. Learning graph-based embedding for personalized product recommendation[J]. Chinese Journal of Computers, 2019, 42(8): 1767-1778.
[4] PARK M J, LEE K G. Exploiting negative preference in content-based music recommendation with contrastive learning[C]//Proceedings of the 16th ACM Conference on Recommender Systems. New York: ACM, 2022: 229-236.
[5] CHANG J X, GAO C, HE X N, et al. Bundle recommendation and generation with graph neural networks[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(3): 2326-2340.
[6] MA Y S, HE Y Z, ZHANG A, et al. CrossCBR: cross-view contrastive learning for bundle recommendation[C]//Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York: ACM, 2022: 1233-1241.
[7] XU F Y, ZHU Z F, LIU P Y, et al. C3BR: category-aware cross-view contrastive learning framework for bundle recommendation[C]//Proceedings of the 28th International Conference on Database Systems for Advanced Applications. Switzerland: Springer Nature, 2023: 194-203.
[8] WANG X, JIN H Y, ZHANG A, et al. Disentangled graph collaborative filtering[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2020: 1001-1010.
[9] ZHAO S, WEI W, ZOU D, et al. Multi-view intent disentangle graph networks for bundle recommendation[C]//Proceedings of the AAAI Conference on Artificial Intelligence. California: AAAI Press, 2022: 4379-4387.
[10] TAN M, CHEN W, WANG W Q, et al. Intention-oriented hierarchical bundle recommendation with preference transfer[C]//Proceedings of the 2021 IEEE International Conference on Web Services. Piscataway: IEEE, 2021: 107-116.
[11] RENDLE S, FREUDENTHALER C, SCHMIDT-THIEME L. Factorizing personalized Markov chains for next-basket recommendation[C]//Proceedings of the 19th International Conference on World Wide Web. New York: ACM, 2010: 811-820.
[12] CHEN L, LIU Y, HE X, et al. Matching user with item set: collaborative bundle recommendation with deep attention network[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence. California: AAAI Press, 2019: 2095-2101
[13] AVNY BROSH T, LIVNE A, SAR SHALOM O, et al. BRUCE: bundle recommendation using contextualized item embeddings[C]//Proceedings of the 16th ACM Conference on Recommender Systems. New York: ACM, 2022: 237-245.
[14] SU C, CHEN M, XIE X Z. Graph convolutional matrix completion via relation reconstruction[C]//Proceedings of the 2021 10th International Conference on Software and Computer Applications. New York: ACM, 2021: 51-56.
[15] WANG X, HE X N, WANG M, et al. Neural graph collaborative filtering[C]//Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2019: 165-174.
[16] CHEN B, XU Y, ZHEN J R, et al. NRMG: news recommendation with multiview graph convolutional networks[J]. IEEE Transactions on Computational Social Systems, 2024, 11(2): 2245-2255.
[17] 吴静, 谢辉, 姜火文. 图神经网络推荐系统综述[J]. 计算机科学与探索, 2022, 16(10): 2249-2263.
WU J, XIE H, JIANG H W. Survey of graph neural network in recommendation system[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(10): 2249-2263.
[18] DENG Q L, WANG K, ZHAO M H, et al. Personalized bundle recommendation in online games[C]//Proceedings of the 29th ACM International Conference on Information & Knowledge Management. New York: ACM, 2020: 2381-2388.
[19] WANG N, SUN J B, LI J B. Cross-level relational graph contrastive learning for bundle recommendation[C]//Proceedings of the 2023 IEEE International Conference on Web Services. Piscataway: IEEE, 2023: 112-117.
[20] KE H L, LI L, WANG P P, et al. Tree-like interaction learning for bundle recommendation[C]//Proceedings of the ICASSP 2023—2023 IEEE International Conference on Acoustics, Speech and Signal Processing. Piscataway: IEEE, 2023: 1-5.
[21] REN Y Y, ZHANG H N, FU L Y, et al. Distillation-enhanced graph masked autoencoders for bundle recommendation[C]//Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2023: 1660-1669.
[22] WEI W, HUANG C, XIA L H, et al. Contrastive meta learning with behavior multiplicity for recommendation[C]//Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining. New York: ACM, 2022: 1120-1128.
[23] YANG H R, CHEN H X, LI L, et al. Hyper meta-path contrastive learning for multi-behavior recommendation[C]//Proceedings of the 2021 IEEE International Conference on Data Mining. Piscataway: IEEE, 2021: 787-796.
[24] WU J H, FAN W Q, CHEN J F, et al. Disentangled contrastive learning for social recommendation[C]//Proceedings of the 31st ACM International Conference on Information & Knowledge Management. New York: ACM, 2022: 4570-4574.
[25] CHEN Y J, LIU Z W, LI J, et al. Intent contrastive learning for sequential recommendation[C]//Proceedings of the ACM Web Conference 2022. New York: ACM, 2022: 2172-2182.
[26] YANG Y H, HUANG C, XIA L H, et al. Debiased contrastive learning for sequential recommendation[C]//Proceedings of the ACM Web Conference 2023. New York: ACM, 2023: 1063-1073.
[27] WU J C, WANG X, FENG F L, et al. Self-supervised graph learning for recommendation[C]//Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2021: 726-735.
[28] 王永贵, 邹赫宇. 多任务联合学习的图卷积神经网络推荐[J]. 计算机工程与应用, 2024, 60(4): 306-314.
WANG Y G, ZOU H Y. Multi-task joint learning for graph convolutional neural network recommendations[J]. Computer Engineering and Applications, 2024, 60(4): 306-314.
[29] GUO J W, HUANG K Z, YI X P, et al. Learning disentangled graph convolutional networks locally and globally[J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35(3): 3640-3651.
[30] 杜晓宇, 陈正, 项欣光. 基于解耦图神经网络的可解释标签感知推荐算法[J]. 软件学报, 2023, 34(12): 5670-5685.
DU X Y, CHEN Z, XIANG X G. Explainable tag-aware recommendation based on disentangled graph neural network[J]. Journal of Software, 2023, 34(12): 5670-5685.
[31] MA J X, ZHOU C, CUI P, et al Learning disentangled representations for recommendation[C]//Advances in Neural Information Processing Systems, 2019: 5712-5723.
[32] MA J X, CUI P, KUANG K, et al. Disentangled graph convolutional networks[C]//Proceedings of the 36th International Conference on Machine Learning. New York: ACM, 2019: 4212-4221.
[33] LI Y Y, HAO Y J, ZHAO P P, et al. Edge-enhanced global disentangled graph neural network for sequential recommendation[J]. ACM Transactions on Knowledge Discovery from Data, 2023, 17(6): 1-22.
[34] CAO J X, LIN X X, CONG X, et al. DisenCDR: learning disentangled representations for cross-domain recommendation[C]//Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2022: 267-277.
[35] STEFFEN R, CHRISTOPH F, ZENO G, et al. BPR: bayesian personalized ranking from implicit feedback[C]//Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence, 2009: 452-461.
[36] CAO D, NIE L Q, HE X N, et al. Embedding factorization models for jointly recommending items and user generated lists[C]//Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2017: 585-594.
[37] CHEN W, HUANG P P, XU J M, et al. POG: personalized outfit generation for fashion recommendation at alibaba iFashion[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York: ACM, 2019: 2662-2670.
[38] YU Z X, LI J T, CHEN L, et al. Unifying multi-associations through hypergraph for bundle recommendation[J]. Knowledge-Based Systems, 2022, 255: 109755.
[39] MA Y S, HE Y Z, WANG X, et al. MultiCBR: multi-view contrastive learning for bundle recommendation[J]. ACM Transactions on Information Systems, 2024, 42(4): 1-23.
[40] HE X N, DENG K, WANG X, et al. LightGCN: simplifying and powering graph convolution network for recommendation[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM, 2020: 639-648.