Page 336 - 《软件学报》2025年第12期

钱忠胜 等: 结合时间间隔数据增强的对偶视图自监督会话推荐模型 5717

5 总结与下一步工作

本文提出一种结合时间间隔数据增强的对偶视图自监督会话推荐模型 TIDA-DSSR, 利用时间间隔数据增强
充实会话内信息, 更合理地建模用户兴趣偏好, 提高推荐准确度; 设计对偶视图编码器, 利用超图卷积网络和
Transformer 编码器从不同角度对用户会话进行建模, 结合时间间隔数据增强模块可更全面、细致地捕捉用户会
话间的隐藏高阶关系, 丰富会话信息多样性; 构建一种新的自监督学习框架, 在对比学习中加入原始会话信息作为
辅助任务, 使得数据增强后的会话信息更全面的同时, 还可减少数据增强早期可能带来无关项目的影响, 提升模型
泛化能力.
所提模型 TIDA-DSSR 无论是与经典模型还是最新模型对比, 在 2 个常用推荐指标上均有明显提升, 但依然
存在一些待完善之处. 在接下来工作中将继续探讨会话中不同关系建模的可能性, 分析这些关系的内在影响规律,
据此挖掘更多影响模型性能的因素, 从而进一步优化模型.

References:
[1] Wang SJ, Pasi G, Hu L, Cao LB. The era of intelligent recommendation: Editorial on intelligent recommendation with advanced AI and
learning. IEEE Intelligent Systems, 2020, 35(5): 3–6. [doi: 10.1109/MIS.2020.3026430]
[2] Hidasi B, Karatzoglou A, Baltrunas L, Tikk D. Session-based recommendations with recurrent neural networks. In: Proc. of the 4th Int’l
Conf. on Learning Representations. San Juan: OpenRiew.net, 2016. 1–10.
[3] Qiu RH, Huang Z, Li JJ, Yin HZ. Exploiting cross-session information for session-based recommendation with graph neural networks.
ACM Trans. on Information Systems (TOIS), 2020, 38(3): 22. [doi: 10.1145/3382764]
[4] Bu JJ, Tan SL, Chen C, Wang C, Wu H, Zhang LJ, He XF. Music recommendation by unified hypergraph: Combining social media
information and music content. In: Proc. of the 18th ACM Int’l Conf. on Multimedia. Firenze: ACM, 2010. 391–400. [doi: 10.1145/
1873951.1874005]
[5] Yadati N, Nimishakavi M, Yadav P, Nitin V, Louis A, Talukdar P. HyperGCN: A new method of training graph convolutional networks
on hypergraphs. In: Proc. of the 33rd Int’l Conf. on Neural Information Processing Systems. Curran Associates Inc., 2019. 135.
[6] Jiang JW, Wei YX, Feng YF, Cao JX, Gao Yue. Dynamic hypergraph neural networks. In: Proc. of the 28th Int’l Joint Conf. on Artificial
Intelligence. Macao: ijcai.org, 2019. 2635–2641. [doi: 10.24963/ijcai.2019/366]
[7] Bandyopadhyay S, Das K, Narasimha Murty M. Line hypergraph convolution network: Applying graph convolution for hypergraphs.
arXiv:2002.03392, 2020.
[8] Xia X, Yin HZ, Yu JL, Wang QY, Cui LZ, Zhang XL. Self-supervised hypergraph convolutional networks for session-based
recommendation. In: Proc. of the 35th AAAI Conf. on Artificial Intelligence. AAAI, 2021. 4503–4511. [doi: 10.1609/aaai.v35i5.16578]
[9] Xie X, Sun F, Liu ZY, Wu SW, Gao JY, Zhang JD, Ding BL, Cui B. Contrastive learning for sequential recommendation. In: Proc. of the
38th IEEE Int’l Conf. on Data Engineering. Kuala Lumpur: IEEE, 2022. 1259–1273. [doi: 10.1109/ICDE53745.2022.00099]
[10] Wang W, Zhang W, Liu SK, Liu Q, Zhang B, Lin LY, Zha HY. Beyond clicks: Modeling multi-relational item graph for session-based
target behavior prediction. In: Proc. of the 2020 Web Conf. Taipei: ACM, 2020. 3056–3062. [doi: 10.1145/3366423.3380077]
[11] Wang ZY, Wei W, Cong G, Li XL, Mao XL, Qiu MH. Global context enhanced graph neural networks for session-based
recommendation. In: Proc of the 43rd Int’l ACM SIGIR Conf. on Research and Development in Information Retrieval. ACM, 2020.
169–178. [doi: 10.1145/3397271.3401142]
[12] Huang ZH, Lin XL, Sun LS, Tang Y, Chen YW. Feature augmentation based graph neural recommendation method in session scenarios.
Chinese Journal of Computers, 2022, 45(4): 766–780 (in Chinese with English abstract). [doi: 10.11897/SP.J.1016.2022.00766]
[13] Zhang S, Gao M, Wen JH, Xiong QY, Tang X. Self-supervised learning for alleviating popularity bias in recommender systems. Acta
Electronica Sinica, 2022, 50(10): 2361–2371 (in Chinese with English abstract). [doi: 10.12263/DZXB.20210443]
[14] Yao TS, Yi XY, Cheng DZ, Yu F, Chen T, Menon A, Hong LC, Chi EH, Tjoa S, Kang JQ, Ettinger E. Self-supervised learning for large-
scale item recommendations. In: Proc. of the 30th ACM Int’l Conf. on Information & Knowledge Management. Queensland: ACM, 2021.
4321–4330. [doi: 10.1145/3459637.3481952]
[15] Zhou K, Wang H, Zhao WX, Zhu YT, Wang SR, Zhang FZ, Wang ZY, Wen JR. S3-Rec: Self-supervised learning for sequential
recommendation with mutual information maximization. In: Proc. of the 29th ACM Int’l Conf. on Information & Knowledge
Management. ACM, 2020. 1893–1902. [doi: 10.1145/3340531.3411954]
[16] Qiu RH, Huang Z, Yin HZ, Wang ZJ. Contrastive learning for representation degeneration problem in sequential recommendation. In: