基于跨会话知识图谱的图注意力网络推荐方法

doi:10.19678/j.issn.1000-3428.0063682

摘要/Abstract

摘要： 基于会话的推荐方法旨在根据匿名用户行为序列预测下一个项目。然而，现有会话推荐方法多基于当前会话建模用户偏好，忽略了会话间蕴含的语义信息及知识图谱中丰富的实体和关系信息，无法有效缓解数据稀疏性的问题。提出一种基于跨会话信息与知识图谱的图注意力网络推荐方法。通过有效整合跨会话信息和知识图谱中的项目知识构建跨会话知识图谱，利用知识感知的注意力机制计算各邻居节点的重要性分数，以更新项目节点表示，采用门控循环单元和图注意力网络将每个会话表示为该会话的当前偏好和全局偏好的组合。在此基础上，将会话嵌入和项目嵌入拼接后输入到多层感知机，得到目标会话和候选项目的预测分数，从而实现会话推荐。实验结果表明，与GRU4REC、SR-GNN、FGNN等方法相比，该方法在KKBOX和JDATA两个真实数据集上的推荐命中率分别至少提高了8.23和2.41个百分点，能有效增强会话推荐性能。

Abstract: The Session-Based Recommendation(SBR) method aims to predict the next item according to the sequence of anonymous user behaviors.However, most existing session recommendation methods typically model user preferences based on the current session, ignoring the semantic information between sessions and the rich relationship and entity information in the knowledge graph, which cannot effectively alleviate the data sparsity problem.This paper presents a recommendation method for a graph attention network based on the cross-session information and knowledge graph.The cross-session knowledge graph is constructed by effectively integrating cross-session information and project knowledge in the knowledge graph.The importance score of each neighbor node is calculated based on the attention mechanism of knowledge perception to update the project node representation.Each session is represented as a combination of the current and global preferences of the session using the Gated Recurrent Unit (GRU) and the graph attention network.Next, session embedding and item embedding are spliced and input into the Multi-Layer Perceptron (MLP) to obtain the prediction scores of target sessions and candidate items to achieve session recommendation.The experimental results show that compared with GRU4REC, SR-GNN, FGNN and other methods, the recommended hit rate of the proposed method on KKBOX and JDATA real data sets has increased by 8.23 and 2.41 percentage points respectively, which can effectively enhance the performance of session recommendation.

Key words: session recommendation, cross-session, knowledge graph, graph attentive network, Gated Recurrent Unit(GRU)

中图分类号:

TP391

张晓晖, 马慧芳, 王文涛, 高子皓. 基于跨会话知识图谱的图注意力网络推荐方法[J]. 计算机工程, 2023, 49(2): 136-142,149.

ZHANG Xiaohui, MA Huifang, WANG Wentao, GAO Zihao. Graph Attention Network Recommendation Method Based on Cross-Session Knowledge Graph[J]. Computer Engineering, 2023, 49(2): 136-142,149.

http://www.ecice06.com/CN/Y2023/V49/I2/136

图/表 9

20230216180915

20230216180918

20230216180921

20230216180924

20230216180928

20230216180931

20230216180934

20230216180938

20230216180941

参考文献

[1] WANG S J, CAO L B, WANG Y, et al.A survey on session-based recommender systems[J].ACM Computing Surveys, 2022, 54(7):154.
[2] ZHOU H C, TAN Q Y, HUANG X, et al.Temporal augmented graph neural networks for session-based recommendations[C]//Proceedings of the 44th International Conference on Research and Development in Information Retrieval.New York, USA:ACM Press, 2021:1798-1802.
[3] 陈恩华, 方宝富.基于物品嵌入向量的会话型推荐算法[J].计算机工程, 2021, 47(7):74-80. CHEN E H, FANG B F.Session-based recommendation algorithm with Item2Vec[J].Computer Engineering, 2021, 47(7):74-80.(in Chinese)
[4] 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.(in Chinese)
[5] 李琳, 唐守廉.基于多层注意力表示的音乐推荐模型[J].电子学报, 2020, 48(9):1672-1679. LI L, TANG S L.Hierarchical attention representation model for music recommendation[J].Acta Electronica Sinica, 2020, 48(9):1672-1679.(in Chinese)
[6] 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, USA:ACM Press, 2010:811-820.
[7] REN P J, CHEN Z M, LI J, et al.RepeatNet:a repeat aware neural recommendation machine for session-based recommendation[C]//Proceedings of the AAAI Conference on Artificial Intelligence.[S.l.]:AAAI Press, 2019:4806-4813.
[8] 高茂庭, 徐彬源.基于循环神经网络的推荐算法[J].计算机工程, 2019, 45(8):198-202, 209. GAO M T, XU B Y.Recommendation algorithm based on recurrent neural network[J].Computer Engineering, 2019, 45(8):198-202, 209.(in Chinese)
[9] HIDASI B, KARATZOGLOU A, BALTRUNAS L, et al.Session-based recommendations with recurrent neural networks[C]//Proceedings of the 4th International Conference on Learning Representation.Washington D.C., USA:[s.n], 2016:1-10.
[10] LI J, REN P J, CHEN Z M, et al.Neural attentive session-based recommendation[C]//Proceedings of Conference on Information and Knowledge Management.New York, USA:ACM Press, 2017:1419-1428.
[11] 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 International Conference on Knowledge Discovery & Data Mining.New York, USA:ACM Press, 2018:1831-1839.
[12] CHOI M J, KIM J H, LEE J, et al.S-walk:accurate and scalable session-based recommendation with random walks[EB/OL].[2021-11-20].https://arxiv.org/abs/2201.01091v1.
[13] JIANG Y B, MA H F, LIU Y H, et al.Enhancing social recommendation via two-level graph attentional networks[J].Neurocomputing, 2021, 449:71-84.
[14] 徐增敏, 陈凯, 郭威伟, 等.面向轻量级卷积网络的激活函数与压缩模型[J].计算机工程, 2022, 48(5):242-250. XU Z M, CHEN K, GUO W W, et al.Activation function and compression model for lightweight convolutional network[J].Computer Engineering, 2022, 48(5):242-250.(in Chinese)
[15] ZHANG X H, MA H F, GAO Z H, et al.Enhancing session-based recommendation with global context information and knowledge graph[C]//Proceedings of the 27 International Conference on Database Systems for Advanced Applications.Berlin, Germany:Springer, 2022:281-288.
[16] LI Y, CHEN T, ZHANG P F, et al.Lightweight self-attentive sequential recommendation[C]//Proceedings of the 30th International Conference on Information & Knowledge Management.New York, USA:ACM Press, 2021:967-977.
[17] WU S, TANG Y Y, ZHU Y Q, et al.Session-based recommendation with graph neural networks[C]//Proceedings of Conference on Artificial Intelligence.[S.l.]:AAAI Press, 2019:346-353.
[18] XU C F, ZHAO P P, LIU Y C, et al.Graph contextualized self-attention network for session-based recommendation[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence.Washington D.C., USA:[s.n.], 2019:3940-3946.
[19] QIU R H, HUANG Z, LI J J, et al.Exploiting cross-session information for session-based recommendation with graph neural networks[J].ACM Transactions on Information Systems, 2020, 38(3):22.
[20] WANG X, HE X N, CAO Y X, et al.KGAT:knowledge graph attention network for recommendation[C]//Proceedings of the 25th International Conference on Knowledge Discovery & Data Mining.New York, USA:ACM Press, 2019:950-958.
[21] LIN Y K, LIU Z Y, SUN M S, et al.Learning entity and relation embeddings for knowledge graph completion[C]//Proceedings of the 29th AAAI Conference on Artificial Intelligence.[S.l.]:AAAI Press, 2015:2181-2187.
[22] QIU J Z, TANG J, MA H, et al.DeepInf:social influence prediction with deep learning[C]//Proceedings of the 24th International Conference on Knowledge Discovery & Data Mining.New York, USA:ACM Press, 2018:2110-2119.
[23] CHANG J X, GAO C, ZHENG Y, et al.Sequential recommendation with graph neural networks[C]//Proceedings of the 44th International Conference on Research and Development in Information Retrieval.New York, USA:ACM Press, 2021:378-387.
[24] GLOROT X, BENGIO Y.Understanding the difficulty of training deep feedforward neural networks[J].Journal of Machine Learning Research, 2010, 9:249-256.
[25] KINGMA D P, BA J L.Adam:a method for stochastic optimization[EB/OL].[2021-11-20].https://arxiv.org/pdf/1412.6980.pdf.

选择文件类型/文献管理软件名称

选择包含的内容