时间感知增强的动态图神经网络序列推荐算法

doi:10.3778/j.issn.1002-8331.2307-0012

摘要/Abstract

摘要： 现有的图神经网络序列推荐方法只关注序列的时序信息而没考虑到序列的时间间隔信息，且只采用单任务的模式进行序列推荐，忽略了可以增强数据，提高泛化能力的辅助任务，这将导致用户的动态交互偏好以及交互时间信息无法被清晰地捕捉。为了缓解上述问题带来的影响，提出了增强时间感知的动态图神经网络序列推荐算法（time-aware enhancement dynamic graph neural networks for sequential recommendation，TaDGSR），具有两方面的优点。该方法将序列构造为一个具有时序信息的动态图，在此基础上融入时间间隔信息并增加时间门控注意力网络模块，以便在捕捉序列间的高阶动态连接的同时增强对时间信息的充分利用。该方法采用了时间阈值分割的长短期预估任务作为其辅助任务，加强时间间隔信息表征的利用，使模型更好地捕捉不同交互间隔的用户动态偏好，最终提升序列推荐任务性能。通过在Amazon电商的Beauty数据集、Games数据集以及CDs数据集上进行的实验，结果表明：（1）与目前较新的基准方法相比，所提方法在三个数据集的Hit@10和NDCG@10指标值上分别取得了4.47%、4.37%、2.72%、1.16%、4.61%、3.97%的平均提升；（2）增加时间间隔信息可以有效提高动态图神经网络的推荐性能，采用时间门控注意力模块以及长短期预估辅助任务对预测性能均能带来正面提升。

关键词: 序列推荐, 图神经网络, 邻域聚合, 时间感知, 多任务学习

Abstract: The existing graph neural network sequence recommendation methods only focus on the temporal information of sequences without considering the time interval information of sequences, and only use single-task mode for sequence recommendation, ignoring the auxiliary tasks that can enhance the data and improve the generalization ability, which will lead to the dynamic interaction preferences of users and interaction time information cannot be captured clearly. To alleviate the impact of these problems, a dynamic graph neural network sequence recommendation algorithm with enhanced time awareness (TaDGSR) is proposed, which has two advantages. Firstly, the method constructs a sequence as a dynamic graph with temporal information, incorporates time interval information and adds a time-gated attention network module to enhance the utilization of temporal information while capturing the higher-order dynamic connections between sequences. Secondly, the method adopts the long short-term prediction task with temporal threshold segmentation as its auxiliary task to enhance the utilization of temporal interval information representation, so that the model can better capture the dynamic preferences of users at different interaction intervals, and finally improve the performance of the sequence recommendation task. Experiments on the Beauty dataset, Games dataset, and CDs dataset of Amazon.com show that：(1) the proposed method achieves average improvements of 4.47%, 4.37%, 2.72%, 1.16%, 4.61% and 3.97% in the Hit@10 and NDCG@10 metric values for the three datasets, respectively, compared to the current newer benchmark method; (2) adding time interval information can effectively improve the recommendation performance of dynamic graph neural networks, and the use of a time-gated attention module and a long- and short-term prediction assistance task can both bring positive improvements to the prediction performance.

Key words: sequential recommendation, graph neural network, neighborhood aggregation, time-aware, multi-tasking learning

陈万志, 王军. 时间感知增强的动态图神经网络序列推荐算法[J]. 计算机工程与应用, 2024, 60(20): 142-152.

CHEN Wanzhi, WANG Jun. Time-Aware Enhancement Dynamic Graph Neural Networks for Sequential Recommendation Algorithm[J]. Computer Engineering and Applications, 2024, 60(20): 142-152.

参考文献

[1] 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, 2010: 811-820.
[2] HIDASI B, KARATZOGLOU A, BALTRUNAS L, et al. Session-based recommendations with recurrent neural networks[J]. arXiv:1511.06939, 2015.
[3] QUADRANA M, KARATZOGLOU A, HIDASI B, et al. Personalizing session-based recommendations with hierarchical recurrent neural networks[C]//Proceedings of the 11th ACM Conference on Recommender Systems, 2017: 130-137.
[4] HIDASI B, KARATZOGLOU A. Recurrent neural networks with top-k gains for session-based recommendations[C]//Proceedings of the 27th ACM International Conference on Information and Knowledge Management, 2018: 843-852.
[5] SAK H, SENIOR A W, BEAUFAYS F. Long short-term memory recurrent neural network architectures for large scale acoustic modeling[C]//Proceedings of the 15th Annual Conference of the International Speech Communication Association, 2014: 41-45.
[6] CHUNG J Y, GULCEHRE C, CHO K H, et al. Empirical evaluation of gated recurrent neural networks on sequence modeling[J]. arXiv:1412.3555, 2014.
[7] TANG J X, WANG K. Personalized top-n sequential recommendation via convolutional sequence embedding[C]//Proceedings of the 11th ACM International Conference on Web Search and Data Mining, 2018: 565-573.
[8] KANG W C, MCAULEY J. Self-attentive sequential recommendation[C]//Proceedings of the 2018 IEEE International Conference on Data Mining, 2018: 197-206.
[9] LIU Q, ZENG Y F, MOKHOSI R, et al. STAMP: short-term attention/memory priority model for session-based recommendation[C]//Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2018: 1831-1839.
[10] KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks[J]. arXiv:1609.02907, 2016.
[11] VELI?KOVI? P, CUCURULL G, CASANOVA A, et al. Graph attention networks[J]. arXiv:1710.10903, 2017.
[12] WU S, TANG Y Y, ZHU Y Q, et al. Session-based recommendation with graph neural networks[C]//Proceedings of the 33rd AAAI Conference on Artificial Intelligence, 2019: 346-353.
[13] ZHANG M Q, WU S, GAO M, et al. Personalized graph neural networks with attention mechanism for session-aware recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2020, 34(8): 3946-3957.
[14] ZHANG M Q, WU S, YU X L, et al. Dynamic graph neural networks for sequential recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(5): 4741-4753.
[15] YANG Z, DING M, XU B, et al. STAM: a spatiotemporal aggregation method for graph neural network-based recommendation[C]//Proceedings of the ACM Web Conference 2022, 2022: 3217-3228.
[16] 陈聪, 张伟, 王骏. 带有时间预测辅助任务的会话式序列推荐[J]. 计算机学报, 2021, 44(9): 1841-1853.
CHEN C, ZHANG W, WANG J. Session-based sequential recommendation with auxiliary time prediction[J]. Chinese Journal of Computers, 2021, 44(9): 1841-1853.
[17] 郭磊, 李秋菊, 刘方爱, 等. 基于自注意力网络的共享账户跨域序列推荐[J]. 计算机研究与发展, 2021, 58(11): 2524-2537.
GUO L, LI Q J, LIU F A, et al. Shared-account cross-domain sequential recommendation with self-attention network[J]. Journal of Computer Research and Development, 2021, 58(11): 2524-2537.
[18] HUANG H, FANG Z X, WANG X, et al. Motif-preserving temporal network embedding[C]//Proceedings of the 29th International Conference on International Joint Conferences on Artificial Intelligence, 2021: 1237-1243.
[19] 任豪, 刘柏嵩, 孙金杨, 等. 基于时间和关系感知的图协同过滤跨域序列推荐[J]. 计算机研究与发展, 2023, 60(1): 112-124.
REN H, LIU B S, SUN J Y, et al. A time and relation-aware graph collaborative filtering for cross-domain sequential recommendation[J]. Journal of Computer Research and Development, 2023, 60(1): 112-124.
[20] 钱忠胜, 杨家秀, 李端明, 等. 结合用户长短期兴趣与事件影响力的事件推荐策略[J]. 计算机研究与发展, 2022, 59(12): 2803-2815.
QIAN Z S, YANG J X, LI D M, et al. Event recommendation strategy combining user long-short term interest and event influence[J]. Journal of Computer Research and Development, 2022, 59(12): 2803-2815.
[21] MA C, MA L H, ZHANG Y X, et al. Memory augmented graph neural networks for sequential recommendation[C]//Proceedings of the 34th AAAI Conference on Artificial Intelligence, 2020: 5045-5052.
[22] LI J C, WANG Y J, MCAULEY J. Time interval aware self-attention for sequential recommendation[C]//Proceedings of the 13th ACM International Conference on Web Search and Data Mining, 2020: 322-330.
[23] 吴博, 梁循, 张树森, 等. 图神经网络前沿进展与应用[J]. 计算机学报, 2022, 45(1): 35-68.
WU B, LIANG X, ZHANG S S, et al. Advances and applications in graph neural network[J]. Chinese Journal of Computers, 2022, 45(1): 35-68.
[24] YING R, HE R N, CHEN K F, et al. Graph convolutional neural networks for web-scale recommender systems[C]//Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2018: 974-983.
[25] 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, 2019: 165-174.
[26] HUANG L W, MA Y T, LIU Y B, et al. Position-enhanced and time-aware graph convolutional network for sequential recommendations[J]. ACM Transactions on Information Systems, 2021, 6: 1-32.
[27] TIAN Y, CHANG J X, NIU Y N, et al. When multi-level meets multi-interest: a multi-grained neural model for sequential recommendation[C]//Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2022: 1632-1641.
[28] XU D, RUAN C W, KORPEOGLU E, et al. Inductive representation learning on temporal graphs[J]. arXiv:2002.07962, 2020.
[29] SANKAR A, WU Y H, GOU L, et al. DySAT: deep neural representation learning on dynamic graphs via self-attention networks[C]//Proceedings of the 13th ACM International Conference on Web Search and Data Mining, 2020: 519-527.
[30] HADASH G, SHALOM O S, OSADCHY R. Rank and rate: multi-task learning for recommender systems[C]//Proceedings of the 12th ACM Conference on Recommender Systems, 2018: 451-454.
[31] MA X, ZHAO L Q, HUANG G, et al. Entire space multi-task model: an effective approach for estimating post-click conversion rate[C]//Proceedings of the 41st International ACM SIGIR Conference on Research & Development in Information Retrieval, 2018: 1137-1140.
[32] YANG C X, PAN J W, GAO X F, et al. Cross-task knowledge distillation in multi-task recommendation[J]. arXiv:2202.
09852, 2022.
[33] MCAULEY J, TARGETT C, SHI Q, et al. Image-based recommendations on styles and substitutes[C]//Proceedings of the 38th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2015: 43-52.
[34] HE X N, LIAO L Z, ZHANG H W, et al. Neural collaborative filtering[C]//Proceedings of the 26th International Conference on World Wide Web, 2017: 173-182.
[35] MA C, KANG P, LIU X. Hierarchical gating networks for sequential recommendation[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2019: 825-833.
[36] WANG J L, DING K Z, HONG L J, et al. Next-item recommendation with sequential hypergraphs[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, 2020: 1101-1110.
[37] CHANG J, GAO C, ZHENG Y, et al. Sequential recommendation with graph neural networks[C]//Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2021: 378-387.
[38] XIE Y, ZHOU P, KIM S. Decoupled side information fusion for sequential recommendation[C]//Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2022: 1611-1621.
[39] KINGMA D P, BA J. Adam: a method for stochastic optimization[J]. arXiv:1412.6980, 2014.