图神经网络推荐系统综述

doi:10.3778/j.issn.1673-9418.2203004

计算机科学与探索 ›› 2022, Vol. 16 ›› Issue (10): 2249-2263.DOI: 10.3778/j.issn.1673-9418.2203004

图神经网络推荐系统综述

吴静, 谢辉⁺(), 姜火文

江西科技师范大学数学与计算机科学学院,南昌 330038

收稿日期:2022-03-01 修回日期:2022-06-07 出版日期:2022-10-01 发布日期:2022-06-15
通讯作者: + E-mail: huixie@aliyun.com
作者简介:吴静（1997—）,女,江西九江人,硕士研究生,CCF学生会员,主要研究方向为推荐系统、图神经网络。
谢辉（1978—）,男,河南信阳人,博士,副教授,CCF会员,主要研究方向为数据挖掘、智能推荐。
姜火文（1974—）,男,江西进贤人,博士,教授,CCF会员,主要研究方向为隐私保护、计算机教育。
基金资助:
国家自然科学基金(71561013);国家自然科学基金(61762044);江西省社会科学研究规划项目(20TQ04);江西省高校人文社会科学研究项目(JC17221);江西省高校人文社会科学研究项目(JD18083);江西省高校人文社会科学研究项目(JC18109);江西省教育厅科技计划项目(GJJ211116);江西省教育厅科技计划项目(GJJ170661)

Survey of Graph Neural Network in Recommendation System

WU Jing, XIE Hui⁺(), JIANG Huowen

School of Mathematics and Computer Science, Jiangxi Science and Technology Normal University, Nanchang 330038, China

Received:2022-03-01 Revised:2022-06-07 Online:2022-10-01 Published:2022-06-15
About author:WU Jing, born in 1997, M.S. candidate, student member of CCF. Her research interests include re-commendation system and graph neural networks.
XIE Hui, born in 1978, Ph.D., associate profes-sor, member of CCF. His research interests in-clude data mining and intelligent recommen-dation.
JIANG Huowen, born in 1974, Ph.D., profes-sor, member of CCF. His research interests inclu-de privacy preservation and computer education.
Supported by:
National Natural Science Foundation of China(71561013);National Natural Science Foundation of China(61762044);Social Science Planning Projects in Jiangxi Province(20TQ04);Fund of Humanities and Social Sciences in Universities of Jiangxi Province(JC17221);Fund of Humanities and Social Sciences in Universities of Jiangxi Province(JD18083);Fund of Humanities and Social Sciences in Universities of Jiangxi Province(JC18109);Project of Science and Technology Plan by Education Department of Jiangxi Province(GJJ211116);Project of Science and Technology Plan by Education Department of Jiangxi Province(GJJ170661)

摘要/Abstract

摘要：

推荐系统（RS）因信息冗杂繁多而诞生。由于数据形式的多样化、复杂化以及数据信息量稀疏性,传统的推荐系统已经不能很好地解决目前的问题。图神经网络（GNN）能从图中对边和节点数据进行特征提取和表示,对处理图结构数据具有先天优势,因此在推荐系统中蓬勃发展。将近年的主要研究成果进行了梳理并加以总结,着重从方法、问题两个角度出发,系统性地综述了图神经网络推荐系统。首先,从方法层面阐述了图卷积网络推荐系统、图注意力网络推荐系统、图自动编码器推荐系统、图生成网络推荐系统、图时空网络推荐系统等五大类的图神经网络推荐系统;接着,从问题相似性出发,归纳出序列推荐问题、社交推荐问题、跨域推荐问题、多行为推荐问题、捆绑推荐问题以及基于会话推荐问题等六大类问题;最后,在对已有方法分析和总结的基础上,指出了目前图神经网络推荐系统研究面临的难点,提出相应的研究问题以及未来研究的方向。

关键词: 图神经网络（GNN）, 推荐系统（RS）, 图卷积网络（GCN）

Abstract:

Recommendation system (RS) was introduced because of a lot of information. Due to the diversity, complexity, and sparseness of data, traditional recommendation system can not solve the current problem well. Graph neural network (GNN) can extract and represent the features from edges and nodes data in the graphs and has inherent advantages in processing the graphs structure data, so it flourishes in recommendation system. This paper sorts out the main references of graph neural network in recommendation system in recent years, focuses on the two perspectives of method and problem, and systematically reviews graph neural network in recommendation system. Firstly, from the method level, five graph neural networks of the recommendation system are elaborated, including the graph convolutional network in the recommendation system, graph attention network in the recommendation system, graph autoencoder in the recommendation system, graph generation network in the recommendation system and graph spatial-temporal network in the recommendation system. Secondly, from the perspective of problem similarity, six major problem types are summarized: sequence recommendation, social recommendation, cross-domain recommendation, multi-behavior recommendation, bundle recommendation, and session-based recommen-dation. Finally, based on the analysis and summary of the existing methods, this paper points out the main difficu-lties in the current research on graph neural network in recommendation system, proposes the corresponding issues that can be investigated, and looks forward to the future research directions on this topic.

Key words: graph neural network (GNN), recommendation system (RS), graph convolution network (GCN)

中图分类号:

TP391

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 Jing, XIE Hui, JIANG Huowen. Survey of Graph Neural Network in Recommendation System[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(10): 2249-2263.

图/表 5

参考文献 83

[1]	于蒙, 何文涛, 周绪川, 等. 推荐系统综述[J]. 计算机应用, 2022, 42(6): 1898-1913. DOI
	YU M, HE W T, ZHOU X C, et al. Review of recommen-dation system[J]. Journal of Computer Applications, 2022, 42(6): 1898-1913.
[2]	陈江美, 张文德. 基于位置社交网络的兴趣点推荐系统研究综述[J]. 计算机科学与探索, 2022, 16(7): 1462-1478. DOI
	CHEN J M, ZHANG W D. Review of point of interest recommendation systems based on location-based social networks[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(7): 1462-1478.
[3]	LIU Z Y, ZHOU J. Introduction to graph neural networks[M]. San Rafael: Morgan & Claypool Publishers, 2020.
[4]	HENAFF M, BRUNA J, LECUN Y. Deep convolutional net-works on graph-structured data[J]. arXiv:1506.05163, 2015.
[5]	BASTINGS J, TITOV I, AZIZ W, et al. Graph convolu-tional encoders for syntax-aware neural machine translation[C]// Proceedings of the 2017 Conference on Empirical Me-thods in Natural Language Processing, Copenhagen, Sep 7-11, 2017. Stroudsburg: ACL, 2017: 1957-1967.
[6]	RHEE S, SEO S, KIM S. Hybridapproach of relation net-work and localized graph convolutional filtering for breast cancer subtype classification[J]. arXiv: 1711.05859, 2017.
[7]	ZHANG Y H, QI P, MANNING C D. Graph convolution over pruned dependency trees improves relation extraction[C]// Proceedings of the 2018 Conference on Empirical Me-thods in Natural Language Processing, Brussels, Oct 31-Nov 4, 2018. Stroudsburg: ACL, 2018: 2205-2215.
[8]	GAO C, ZHENG Y, LI N, et al. Graph neural networks for recommender systems: challenges, methods, and directions[J]. arXiv:2109.12843, 2021.
[9]	吴国栋, 查志康, 涂立静, 等. 图神经网络推荐研究进展[J]. 智能系统学报, 2020, 15(1): 14-24.
	WU G D, ZHA Z K, TU L J, et al. Research advances in graph neural network recommendation[J]. Journal of Intelli-gent Systems, 2020, 15(1): 14-24.
[10]	徐冰冰, 岑科廷, 黄俊杰, 等. 图卷积神经网络综述[J]. 计算机学报, 2020, 43(5): 755-780.
	XU B B, CEN K T, HUANG J J, et al. A survey on graph convolutional neural network[J]. Chinese Journal of Com-puters, 2020, 43(5): 755-780.
[11]	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 Con-ference on Knowledge Discovery and Data Mining, London, Aug 19-23, 2018. New York: ACM, 2018: 974-983.
[12]	CHEN J X, LIN G Q, CHEN J X, et al. Towards efficient allocation of graph convolutional networks on hybrid com-putation-in-memory architecture[J]. Science China Informa-tion Sciences, 2021, 64(6): 108-121.
[13]	TRAN D H, SHENG Q Z, ZHANG W E, et al. HeteGraph: graph learning in recommender systems via graph convolu-tional networks[J]. Neural Computing and Applications, 2021. DOI: 10.1007/s00521-020-05667-z. DOI
[14]	SHAFQAT W, BYUN Y C. Incorporating similarity mea-sures to optimize graph convolutional neural networks for product recommendation[J]. Applied Sciences-Basel, 2021, 11(4): 1366.
[15]	YIN Y, ZHENG W. An efficient recommendation algorithm based on heterogeneous information network[J]. Complexity, 2021(17): 1-18.
[16]	CHEN J Y, YU J, LU W J, et al. IR-Rec: an interpretive rules-guided recommendation over knowledge graph[J]. In-formation Sciences, 2021, 563: 326-341.
[17]	BONET E R, NGUYEN D M, DELIGIANNIS N, et al. Temporal collaborative filtering with graph convolutional neural networks[C]// Proceedings of the 25th International Conference on Pattern Recognition, Milan, Jan 10-15, 2021. Piscataway: IEEE, 2021: 4736-4742.
[18]	SONG W P, XIAO Z P, WANG Y F, et al. Session-based social recommendation via dynamic graph attention net-works[C]// Proceedings of the 12th ACM International Con-ference on Web Search and Data Mining, Melbourne, Feb 11-15, 2019. New York: ACM, 2019: 555-563.
[19]	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. DOI URL
[20]	WU Q T, ZHANG H R, GAO X F, et al. Dual graph atten-tion networks for deep latent representation of multifaceted social effects in recommender systems[C]// Proceedings of the Web Conference 2019:Proceedings of the World Wide Web Conference,San Francisco, May 13-17, 2019. New York: ACM, 2019: 2091-2102.
[21]	XIAO Y, PEI Q Q, XIAO T T, et al. MutualRec: joint friend and item recommendations with mutualistic attentional graph neural networks[J]. Journal of Network and Com-puter Applications, 2021, 177: 102954.
[22]	DANG D P, CHEN C X, LI H C, et al. Deep knowledge-aware framework for web service recommendation[J]. The Journal of Supercomputing, 2021, 77: 14280-14304. DOI URL
[23]	LI A C, YANG B. GSIRec: learning with graph side infor-mation for recommendation[J]. World Wide Web-Internet and Web Information Systems, 2021, 24(5): 1411-1437.
[24]	SALAMAT A, LUO X, JAFARI A. HeteroGraphRec: a heterogeneous graph-based neural networks for social recom-mendations[J]. Knowledge-Based Systems, 2021, 217: 106817. DOI URL
[25]	SANG L, XU M, QIAN S S, et al. Knowledge graph en-hanced neural collaborative filtering with residual recurrent network[J]. Neurocomputing, 2021, 454: 417-429. DOI URL
[26]	BEHROUZI T, HATZINAKOS D. Graph variational auto-encoder for deriving EEG-based graph embedding[J]. Pat-tern Recognition, 2022, 121: 108202.
[27]	ZHANG X B, YANG Y, ZHAI D H, et al. Local2Global: unsupervised multi-view deep graph representation lear-ning with nearest neighbor constraint[J]. Knowledge-Based Systems, 2021, 231: 107439. DOI URL
[28]	TRAN Q M, NGUYEN H D, HUYNH T, et al. Measuring the influence and amplification of users on social network with unsupervised behaviors learning and efficient interaction-based knowledge graph[J]. Journal of Combinatorial Opti-mization, 2021. DOI: 10.1007/s10878-021-00815-0. DOI
[29]	ZHENG Q Q, LIU G F, LIU A, et al. Implicit relation-aware social recommendation with variational auto-encoder[J]. World Wide Web-Internet and Web Information Systems, 2021, 24(5): 1395-1410.
[30]	YAO L Y, ZHONG J B, ZHANG X F, et al. Correlated Wasserstein autoencoder for implicit data recommendation[C]// Proceedings of the 2020 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, Melbourne, Dec 14-17, 2020. Piscataway: IEEE, 2020: 417-422.
[31]	DENG K, HUANG J J, QIN J. Hybrid GNN-SR: combi-ning unsupervised and supervised graph learning for ses-sion-based recommendation[C]// Proceedings of the 20th IEEE International Conference on Data Mining Workshops, Sorrento, Nov 17-20, 2020. Piscataway: IEEE, 2020: 136-143.
[32]	OHTOMO K, HARAKAWA R, QGAWA T, et al. Persona-lized recommendation of Tumblr posts using graph convo-lutional networks with preference-aware multimodal fea-tures[J]. Item Transactions on Media Technology and App-lications, 2021, 9(1): 54-61.
[33]	BONGINI P, BIANCHINI M, SCARSELLI F. Molecular generative graph neural networks for drug discovery[J]. Neurocomputing, 2021, 450: 242-252. DOI URL
[34]	ZHOU Y D, DING Z H, LIU X M, et al. Infer-AVAE: an attribute inference model based on adversarial variational autoencoder[J]. Neurocomputing, 2022, 483: 105-115. DOI URL
[35]	XU D, RUAN C W, MOTWANI K, et al. Generative graph convolutional network for growing graphs[C]// Proceedings of the 2019 IEEE International Conference on Acoustics, Speech and Signal Processing, Brighton, May 12-17, 2019. Piscataway: IEEE, 2019: 3167-3171.
[36]	WU F, GAO M, YU J L, et al. Ready for emerging threats to recommender systems? A graph convolution-based gene-rative shilling attack[J]. Information Sciences, 2021, 578: 683-701. DOI URL
[37]	ZHANG S S, NI J C, HOU L J, et al. Global-affine and local-specific generative adversarial network for semantic-guided image generation[J]. Mathematical Foundations of Computing, 2021, 4(3): 145-165. DOI URL
[38]	XU X R, CHEN L M, ZU S P, et al. Hulu video recommen-dation: from relevance to reasoning[C]// Proceedings of the 12th ACM Conference on Recommender Systems, Vancou-ver, Oct 2-7, 2018. New York: ACM, 2018: 482.
[39]	PARK S, JANG D K, LEE S H. Diverse motion stylization for multiple style domains via spatial-temporal graph-based generative model[J]. Proceedings of the ACM on Compu-ter Graphics and Interactive Techniques, 2021, 4(3): 36.
[40]	ZHANG Z, BU J J, LI Z, et al. Tige CMN: on exploration of temporal interaction graph embedding via coupled me-mory neural networks[J]. Neural Networks, 2021, 140: 13-26. DOI URL
[41]	杨珍, 丁铭, 唐杰, 等. 面向推荐系统的时空图卷积方法和系统: CN113269603A[P]. 2021-08-17.
	YANG Z, DING M, TANG J, et al. Temporal and spatial graph convolution method and system for recommendation system: CN113269603A[P]. 2021-08-17.
[42]	HAN H Y, ZHANG M D, HOU M, et al. STGCN: a spatial-temporal aware graph learning method for POI recommen-dation[C]// Proceedings of the 20th IEEE International Con-ference on Data Mining, Sorrento, Nov 17-20, 2020. Pisca-taway: IEEE, 2020: 1052-1057.
[43]	WU S, SUN F, ZHANG W, et al. Graph neural networks in recommender systems: a survey[J]. arXiv:2011.02260, 2020.
[44]	XIANG N H, ZHAO C R, EMINE Y, et al. Graph techno-logies for user modeling and recommendation: introduc-tion to the special issue-part 1[J]. ACM Transactions on In-formation Systems, 2021, 40(2): 1-5.
[45]	YANG M H, CAO S S, HU B B, et al. IntelliTag: an intel-ligent cloud customer service system based on tag recom-mendation[C]// Proceedings of the 2021 IEEE 37th Interna-tional Conference on Data Engineering, Chania, Apr 19-22, 2021. Piscataway: IEEE, 2021: 2559-2570.
[46]	GU P, HAN Y Q, GAO W, et al. Enhancing session-based social recommendation through item graph embedding and contextual friendship modeling[J]. Neurocomputing, 2021, 419: 190-202. DOI URL
[47]	TAO Y, WANG C, YAO L N, et al. Item trend learning for sequential recommendation system using gated graph neu-ral network[J]. Neural Computing & Applications, 2021. DOI: 10.1007/S00521-021-05723-2. DOI
[48]	WANG X, HE X, CAO Y, et al. KGAT: knowledge graph attention network for recommendation[C]// Proceedings of the 25th ACM SIGKDD International Conference on Know-ledge Discovery and Data Mining, Anchorage, Aug 4-8, 2019. New York: ACM, 2019: 950-958.
[49]	GUO Z W, WANG H. A deep graph neural network-based mechanism for social recommendations[J]. IEEE Transac-tions on Industrial Informatics, 2021, 17(4): 2776-2783.
[50]	LIU W W, ZHANG Y, WANG J L, et al. Item relationship graph neural networks for e-commerce[J]. IEEE Transac-tions on Neural Networks and Learning Systems, 2021. DOI: 10.1109/TNNLS.2021.3060872. DOI
[51]	SALAMAT A, LUO X, JAFARI A. Hetero Graph Rec: a heterogeneous graph-based neural networks for social re-commendations[J]. Knowledge-based Systems, 2021, 217: 106817. DOI URL
[52]	LIU Y, YANG S, XU Y, et al. Contextualized graph atten-tion network for recommendation with item knowledge graph[J]. IEEE Transactions on Knowledge and Data En-gineering, 2021. DOI: 10.1109/TKDE.2021.3082948. DOI
[53]	TU K, CUI P, WANG D, et al. Conditional graph attention networks for distilling and refining knowledge graphs in recommendation[C]// Proceedings of the 30th ACM Interna-tional Conference on Information & Knowledge Manage-ment. New York: ACM, 2021: 1834-1843.
[54]	WANG H Y, YAO K M, LUO J, et al. An implicit prefe-rence-aware sequential recommendation method based on knowledge graph[J]. Wireless Communications & Mobile Computing, 2021: 5206228.
[55]	YANG X, HUAN Z Y, ZHAI Y S, et al. Research of per-sonalized recommendation technology based on knowledge graphs[J]. Applied Sciences-Basel, 2021, 11(15): 7104.
[56]	LOANNIDIS V N, ZAMZAM A S, GIANNAKIS G B, et al. Coupled graphs and tensor factorization for recommen-der systems and community detection[J]. IEEE Transactions on Knowledge and Data Engineering, 2021, 33(3): 909-920.
[57]	OUYANG Y, GUO B, TANG X, et al. Mobile App cross-domain recommendation with multi-graph neural network[J]. ACM Transactions on Knowledge Discovery from Data, 2021, 15(4): 55.
[58]	SHEU H S, CHU Z X, QI D Q, et al. Knowledge-guided article embedding refinement for session-based news re-commendation[J]. IEEE Transactions on Neural Networks and Learning Systems, 2021. DOI: 10.1109/TNNLS.2021. 3084958. DOI
[59]	LIANG T T, SHENG X, ZHOU L, et al. Mobile app recom-mendation via heterogeneous graph neural network in edge computing[J]. Applied Soft Computing, 2021, 103: 107162. DOI URL
[60]	MA M Y, NA S, WANG H Y, et al. The graph-based beha-vior-aware recommendation for interactive news[J]. arXiv:1812.00002V2, 2018.
[61]	WANG X, HE X, WANG M, et al. Neural graph colla-borative filtering[C]// Proceedings of the 42nd International ACM SIGIR Conference, Paris, Jul 21-25, 2019. New York: ACM, 2019: 165-174.
[62]	HE X, DENG K, WANG X, et al. Light GCN: simplifying and powering graph convolution network for recommen-dation[C]// Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Infor-mation Retrieval. New York: ACM, 2020: 639-648.
[63]	AMAR B. Revisiting SVD to generate powerful node em-beddings for recommendation systems[J]. arXiv:2110.03665, 2021.
[64]	LIU H D, YANG B, LI D S. Graph collaborative filtering based on dual-message propagation mechanism[J]. IEEE Transactions on Cybernetics, 2021. DOI: 10.1109/TCYB.2021. 3100521. DOI
[65]	XIA L H, HUANG C, XU Y, et al. Multi-behavior enhanced recommendation with cross-interaction collaborative rela-tion modeling[C]// Proceedings of the 2021 IEEE 37th Inter-national Conference on Data Engineering, Chania, Apr 19-22, 2021. Piscataway: IEEE, 2021: 1931-1936.
[66]	YU B, ZHANG R Q, CHEN W, et al. Graph neural network based model for multi-behavior session-based recommen-dation[J]. Geoinformatica, 2022, 26(2): 429-447. DOI URL
[67]	YANG J, MA W, ZHANG M, et al. Legal GNN: legal infor-mation enhanced graph neural network for recommendation[J]. ACM Transactions on Information Systems, 2021, 40(2): 1-29.
[68]	ZHANG C, WANG Y, ZHU L, et al. Multi-graph hetero-geneous interaction fusion for social recommendation[J]. ACM Transactions on Information Systems, 2021, 40(2): 1-26.
[69]	YUAN Y J, WEN D J, WEI Z, et al. A kg-enhanced multi-graph neural network for attentive herb recommendation[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2021. DOI: 10.1109/TCBB.2021.3115489. DOI
[70]	LIU X, SUN Y B, LIU Z W, et al. Learning diverse fashion collocation by neural graph filtering[J]. IEEE Transactions on Multimedia, 2021, 23: 2894-2901. DOI URL
[71]	CHEN J G, LI K L, LI K Q, et al. Dynamic planning of bicycle stations in dock less public bicycle-sharing system using gated graph neural network[J]. ACM Transactions on Intelligent Systems and Technology, 2021, 12(2): 25.
[72]	GONG W J, YU Q S. A deep music recommendation method based on human motion analysis[J]. IEEE Access, 2021, 9: 26290-26300. DOI URL
[73]	LING C Y, ZON Y Z, XIE B. Graph neural network based collaborative filtering for API usage recommendation[C]// Proceedings of the 2021 IEEE International Conference on Software Analysis, Evolution and Reengineering, Honolulu, Mar 9-12, 2021. Piscataway: IEEE, 2021: 36-47.
[74]	ZHANG Y Q, YANG H Y, KUANG L. A web API recom-mendation method with composition relationship based on GCN[C]// Proceedings of the 2020 IEEE International Con-ference on Parallel & Distributed Processing with Appli-cations, Big Data & Cloud Computing, Sustainable Com-puting & Communications, Social Computing & Networking, Exeter, Dec 17-19, 2020. Piscataway: IEEE, 2020: 601-608.
[75]	ZHU Y F, LU H, QIU P, et al. Heterogeneous teaching eva-luation network based offline course recommendation with graph learning and tensor factorization[J]. Neurocomputing, 2021, 415: 84-95. DOI URL
[76]	ZHENG Y J, LIU S Y, LI Z K, et al. DGTN: dual-channel graph transition network for session-based recommendation[C]// Proceedings of the 20th IEEE International Conferen-ce on Data Mining, Sorrento, Nov 17-20, 2020. Piscata-way: IEEE, 2020: 236-242.
[77]	HUANG C, CHEN J, XIA L, et al. Graph-enhanced multi-task learning of multi-level transition dynamics for session-based recommendation[C]// Proceedings of the 2021 AAAI Conference on Artificial Intelligence. Menlo Park: AAAI, 2021: 4123-4130.
[78]	WU Z T, SONG C Y, CHEN Y Q, et al. A review of recom-mendation system research based on bipartite graph[C]// Proceedings of the 2020 2nd International Conference on Computer Science Communication and Network Security, Sanya, Dec 22-23, 2020: 336.
[79]	PAN Z Q, CHEN H H. Collaborative knowledge-enhanced recommendation with self-supervisions[J]. Mathematics, 2021, 9(17): 2129. DOI URL
[80]	GUO J, ZHOU Y, ZHANG P, et al. Trust-aware recommen-dation based on heterogeneous multi-relational graphs fu-sion[J]. Information Fusion, 2021, 74: 87-95. DOI URL
[81]	CHEN W, REN P, CAI F, et al. Improving end-to-end se-quential recommendations with intent-aware diversification[C]// Proceedings of the 29th ACM International Conference on Information & Knowledge Management. New York: ACM, 2020: 175-184.
[82]	ISUFI E, POCCHIARI M, HANJALIC A. Accuracy-diversity trade-off in recommender systems via graph convolutions[J]. Information Processing & Management, 2021, 58(2): 102459. DOI URL
[83]	PAN Z Q, CHEN H H. Efficient graph collaborative filte-ring via contrastive learning[J]. Sensors, 2021, 21(14): 4666. DOI URL

分类	作者	关键技术	问题场景	优点	局限性
图卷积网络推荐系统	Ying等^[11]	GCN、随机游走	Web推荐任务	提高模型的鲁棒性	不能解决其他大规模的图表示学习问题
	Chen等^[12]	GCN	所有推荐任务	减少处理延迟	内存访问模型复杂
	Tran等^[13]	GCN	应用于大规模异构图数据的推荐任务	处理异构图数据	仅适用于两个实体,即用户和项目
	Shafqat等^[14]	GCN	在线产品推荐任务	简化了GCN模型的邻居抽样任务,提高了训练效率,降低了复杂度和计算时间	需要形成会话图,并不适应于所有推荐系统场景
	Yin等^[15]	GCN	异构信息网络的推荐任务	提取和组合异构图中的结构特征,减小了训练规模,提高了计算效率	算法复杂
	Chen等^[16]	GCN、KG	TOP-K推荐	提高可解释性	学习效率低,未利用更多的辅助信息
	Bonet等^[17]	GCN、递归神经网络	大数据推荐任务	提高推荐系统的性能和推荐的准确度	处理不了冷启动和数据稀疏性问题,忽略了推荐系统的可解释性
图注意力网络推荐系统	Song等^[18]	图注意力神经网络	在线社区社交推荐	能进行用户的动态的兴趣推荐	只能对大规模数据有效
	Jiang等^[19]	图注意力神经网络、GCN	社交推荐	能发现潜在的社会传播效应	模型复杂,无法区分社交的正面和负面影响
	Wu等^[20]	图注意力神经网络	社交推荐	能学习社会深层次表征,提高推荐准确度	需要提取足够多的高层联系信息
	Xiao等^[21]	图注意力神经网络	社交推荐	融合用户偏好和社交交互信息	不能完全利用辅助信息
	Dang等^[22]	图注意力神经网络、知识图谱	Web服务	充分挖掘文本特征,解决数据稀疏性问题,优化特征表示,提高推荐的可解释性	模型需与其他开放知识库相结合
	Li等^[23]	图注意力神经网络、知识图谱	评级预测任务、TOP-K推荐任务	解决数据稀疏和冷启动的问题	运行时间较长
	Salamat等^[24]	图注意力神经网络	社交推荐	提高了模型的可解释性	未考虑社交网络的动态行为
	Sang等^[25]	图注意力神经网络、知识图谱、残差递归神经网络	所有推荐	能自动捕捉丰富的语义信息和用户与项目之间复杂的隐含关系	未考虑用户之间交互的顺序性
图自动编码器推荐系统	Zheng等^[29]	图自动编码器、GCN	社交推荐	捕捉隐藏在图结构下的隐式高阶关系,提高推荐系统性能	未考虑用户之间交互的顺序性
	Yao等^[30]	图自动编码器、GCN	隐式数据的推荐系统	捕获数据相关性以提高推荐性能	未考虑时间顺序因素
	Deng等^[31]	图自动编码器、无监督学习、有监督学习	会话推荐	考虑了会话中的项目之间依赖关系	对模型中各组件和超参数的影响未知
	Ohtomo等^[32]	图自动编码器	个性化推荐	从大量帖子中为每个用户个性化推荐帖子	训练时间长
图生成网络推荐系统	Zhou等^[34]	图生成神经网络、GCN	个性化推荐	更好地利用辅助信息并生成不受限制的输出表示	对稀疏性数据很容易产生过拟合
	Xu等^[35]	图生成神经网络、GCN	社交推荐	解决冷启动问题	大图的计算复杂度高
	Wu等^[36]	图生成神经网络、生成对抗网络	在线推荐	增强推荐系统的稳定性	只能对评论等基于内容的推荐有用
	Zhang等^[37]	图生成神经网络、GCN	图像推荐	解决了合成细粒度纹理和小规模实例的困难	严重依赖于推断的语义
	Xu等^[38]	图生成神经网络、GCN	在线视频推荐	提高推荐的准确度	大图长序列建模困难,信息质量要求高
图时空网络推荐系统	Park等^[39]	图时空神经网络、GCN	运动风格推荐	能提取空间和时间两个维度的特征	对随机噪音十分敏感,适合少量已经标好明确样式标签的数据
	Zhang等^[40]	图时空神经网络、图嵌入、GCN	所有推荐任务	适用性广	仅仅考虑了二部图,未扩展到多部异构图而且训练过程中提取数据是均匀抽样,其实用性较差
	杨珍等^[41]	图时空神经网络、GCN	用户商品推荐	提高了推荐系统的性能	只能用于购物商品推荐
	Han等^[42]	图时空神经网络、GCN	POI推荐	缓解数据稀疏性问题	未能考虑到时空序列节点之间的上下文信息

分类	作者	关键技术	问题场景	优点	局限性
图卷积网络推荐系统	Ying等^[11]	GCN、随机游走	Web推荐任务	提高模型的鲁棒性	不能解决其他大规模的图表示学习问题
	Chen等^[12]	GCN	所有推荐任务	减少处理延迟	内存访问模型复杂
	Tran等^[13]	GCN	应用于大规模异构图数据的推荐任务	处理异构图数据	仅适用于两个实体,即用户和项目
	Shafqat等^[14]	GCN	在线产品推荐任务	简化了GCN模型的邻居抽样任务,提高了训练效率,降低了复杂度和计算时间	需要形成会话图,并不适应于所有推荐系统场景
	Yin等^[15]	GCN	异构信息网络的推荐任务	提取和组合异构图中的结构特征,减小了训练规模,提高了计算效率	算法复杂
	Chen等^[16]	GCN、KG	TOP-K推荐	提高可解释性	学习效率低,未利用更多的辅助信息
	Bonet等^[17]	GCN、递归神经网络	大数据推荐任务	提高推荐系统的性能和推荐的准确度	处理不了冷启动和数据稀疏性问题,忽略了推荐系统的可解释性
图注意力网络推荐系统	Song等^[18]	图注意力神经网络	在线社区社交推荐	能进行用户的动态的兴趣推荐	只能对大规模数据有效
	Jiang等^[19]	图注意力神经网络、GCN	社交推荐	能发现潜在的社会传播效应	模型复杂,无法区分社交的正面和负面影响
	Wu等^[20]	图注意力神经网络	社交推荐	能学习社会深层次表征,提高推荐准确度	需要提取足够多的高层联系信息
	Xiao等^[21]	图注意力神经网络	社交推荐	融合用户偏好和社交交互信息	不能完全利用辅助信息
	Dang等^[22]	图注意力神经网络、知识图谱	Web服务	充分挖掘文本特征,解决数据稀疏性问题,优化特征表示,提高推荐的可解释性	模型需与其他开放知识库相结合
	Li等^[23]	图注意力神经网络、知识图谱	评级预测任务、TOP-K推荐任务	解决数据稀疏和冷启动的问题	运行时间较长
	Salamat等^[24]	图注意力神经网络	社交推荐	提高了模型的可解释性	未考虑社交网络的动态行为
	Sang等^[25]	图注意力神经网络、知识图谱、残差递归神经网络	所有推荐	能自动捕捉丰富的语义信息和用户与项目之间复杂的隐含关系	未考虑用户之间交互的顺序性
图自动编码器推荐系统	Zheng等^[29]	图自动编码器、GCN	社交推荐	捕捉隐藏在图结构下的隐式高阶关系,提高推荐系统性能	未考虑用户之间交互的顺序性
	Yao等^[30]	图自动编码器、GCN	隐式数据的推荐系统	捕获数据相关性以提高推荐性能	未考虑时间顺序因素
	Deng等^[31]	图自动编码器、无监督学习、有监督学习	会话推荐	考虑了会话中的项目之间依赖关系	对模型中各组件和超参数的影响未知
	Ohtomo等^[32]	图自动编码器	个性化推荐	从大量帖子中为每个用户个性化推荐帖子	训练时间长
图生成网络推荐系统	Zhou等^[34]	图生成神经网络、GCN	个性化推荐	更好地利用辅助信息并生成不受限制的输出表示	对稀疏性数据很容易产生过拟合
	Xu等^[35]	图生成神经网络、GCN	社交推荐	解决冷启动问题	大图的计算复杂度高
	Wu等^[36]	图生成神经网络、生成对抗网络	在线推荐	增强推荐系统的稳定性	只能对评论等基于内容的推荐有用
	Zhang等^[37]	图生成神经网络、GCN	图像推荐	解决了合成细粒度纹理和小规模实例的困难	严重依赖于推断的语义
	Xu等^[38]	图生成神经网络、GCN	在线视频推荐	提高推荐的准确度	大图长序列建模困难,信息质量要求高
图时空网络推荐系统	Park等^[39]	图时空神经网络、GCN	运动风格推荐	能提取空间和时间两个维度的特征	对随机噪音十分敏感,适合少量已经标好明确样式标签的数据
	Zhang等^[40]	图时空神经网络、图嵌入、GCN	所有推荐任务	适用性广	仅仅考虑了二部图,未扩展到多部异构图而且训练过程中提取数据是均匀抽样,其实用性较差
	杨珍等^[41]	图时空神经网络、GCN	用户商品推荐	提高了推荐系统的性能	只能用于购物商品推荐
	Han等^[42]	图时空神经网络、GCN	POI推荐	缓解数据稀疏性问题	未能考虑到时空序列节点之间的上下文信息

问题分类	方法分类	作者	难点
序列推荐问题	图注意力网络、图卷积网络	Yang等^[45]	数据稀疏和冷启动问题,异构图
	图卷积网络、图注意力网络	Gu等^[46]	动态兴趣建模问题
	图注意力网络	Tao等^[47]	项目趋势信息,动态图构建问题
	图注意力网络	Wang等^[48]	高阶关系建模,可解释性
社交推荐问题	图自动编码器	Guo等^[49]	大数据与个性化信息
	图神经网络	Liu等^[50]	大数据,关系动态变化问题
	图注意力网络	Salamat等^[51]	大数据,异构图,可解释性,动态行为
	图注意力网络	Liu等^[52]	动态表示问题,知识图
	图注意力网络	Tu等^[53]	数据稀疏,个性化问题,知识图
	图卷积网络	Wang等^[54]	大数据,隐含兴趣,动态兴趣
跨域推荐问题	图神经网络	Yang等^[55]	大数据,数据稀疏和冷启动,动态问题
	图神经网络	Loannidis等^[56]	可解释性
	图神经网络	Ouyang等^[57]	数据稀疏
	图注意力网络	Sheu等^[58]	缺乏用户交互记录
	图神经网络	Liang等^[59]	信息高效性,异构图
	图卷积网络、图注意力网络	Ma等^[60]	异构图,多样性和准确性
	图卷积网络	Wang等^[61]	交互图嵌入特征表示
	图卷积网络	He等^[62]	邻域聚合
	图神经网络	Amar^[63]	算法简洁性,信息高效性
	图神经网络	Liu等^[64]	模型精确性
多行为推荐问题	图神经网络	Xia等^[65]	提取多类型下的异构关系
	图神经网络	Yu等^[66]	有效捕获信息
	图卷积网络、图注意力网络	Ma等^[60]	异构图,多样性和准确性
捆绑推荐问题	图神经网络	Yang等^[67]	信息增强问题
	图神经网络	Zhang等^[68]	异构图
	图注意力网络	Yuan等^[69]	异构图
	图神经网络	Liu等^[70]	个性多样化
	图神经网络	Chen等^[71]	动态化,准确性
	图注意力网络、图卷积网络	Yang等^[45]	数据稀疏和冷启动问题,异构图
	图卷积网络	Gong等^[72]	结合深度学习从舞蹈动作中推荐音乐
	图神经网络	Ling等^[73]	信息的高阶连通性
	图卷积网络、图自动编码器	Zhang等^[74]	大数据,数据稀疏
	图神经网络	Zhu等^[75]	数据稀疏和冷启动问题
会话推荐问题	图神经网络	Zheng等^[76]	异构图,潜在信息
	图神经网络	Yu等^[66]	有效捕获信息
	图卷积网络、图注意力网络	Gu等^[46]	动态兴趣建模问题
	图神经网络	Huang等^[77]	动态信息及信息增强

问题分类	方法分类	作者	难点
序列推荐问题	图注意力网络、图卷积网络	Yang等^[45]	数据稀疏和冷启动问题,异构图
	图卷积网络、图注意力网络	Gu等^[46]	动态兴趣建模问题
	图注意力网络	Tao等^[47]	项目趋势信息,动态图构建问题
	图注意力网络	Wang等^[48]	高阶关系建模,可解释性
社交推荐问题	图自动编码器	Guo等^[49]	大数据与个性化信息
	图神经网络	Liu等^[50]	大数据,关系动态变化问题
	图注意力网络	Salamat等^[51]	大数据,异构图,可解释性,动态行为
	图注意力网络	Liu等^[52]	动态表示问题,知识图
	图注意力网络	Tu等^[53]	数据稀疏,个性化问题,知识图
	图卷积网络	Wang等^[54]	大数据,隐含兴趣,动态兴趣
跨域推荐问题	图神经网络	Yang等^[55]	大数据,数据稀疏和冷启动,动态问题
	图神经网络	Loannidis等^[56]	可解释性
	图神经网络	Ouyang等^[57]	数据稀疏
	图注意力网络	Sheu等^[58]	缺乏用户交互记录
	图神经网络	Liang等^[59]	信息高效性,异构图
	图卷积网络、图注意力网络	Ma等^[60]	异构图,多样性和准确性
	图卷积网络	Wang等^[61]	交互图嵌入特征表示
	图卷积网络	He等^[62]	邻域聚合
	图神经网络	Amar^[63]	算法简洁性,信息高效性
	图神经网络	Liu等^[64]	模型精确性
多行为推荐问题	图神经网络	Xia等^[65]	提取多类型下的异构关系
	图神经网络	Yu等^[66]	有效捕获信息
	图卷积网络、图注意力网络	Ma等^[60]	异构图,多样性和准确性
捆绑推荐问题	图神经网络	Yang等^[67]	信息增强问题
	图神经网络	Zhang等^[68]	异构图
	图注意力网络	Yuan等^[69]	异构图
	图神经网络	Liu等^[70]	个性多样化
	图神经网络	Chen等^[71]	动态化,准确性
	图注意力网络、图卷积网络	Yang等^[45]	数据稀疏和冷启动问题,异构图
	图卷积网络	Gong等^[72]	结合深度学习从舞蹈动作中推荐音乐
	图神经网络	Ling等^[73]	信息的高阶连通性
	图卷积网络、图自动编码器	Zhang等^[74]	大数据,数据稀疏
	图神经网络	Zhu等^[75]	数据稀疏和冷启动问题
会话推荐问题	图神经网络	Zheng等^[76]	异构图,潜在信息
	图神经网络	Yu等^[66]	有效捕获信息
	图卷积网络、图注意力网络	Gu等^[46]	动态兴趣建模问题
	图神经网络	Huang等^[77]	动态信息及信息增强

图神经网络推荐系统综述

Survey of Graph Neural Network in Recommendation System

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 5

参考文献 83

相关文章 5

编辑推荐

Metrics

[1]	蒋光峰, 胡鹏程, 叶桦, 仰燕兰. 基于重构误差的同构图分类模型[J]. 计算机科学与探索, 2022, 16(1): 185-193.
[2]	袁立宁, 李欣, 王晓冬, 刘钊. 图嵌入模型综述[J]. 计算机科学与探索, 2022, 16(1): 59-87.
[3]	王晓东, 赵一宁, 肖海力, 王小宁, 迟学斌. 使用GNN与RNN实现用户行为分析[J]. 计算机科学与探索, 2021, 15(5): 838-847.
[4]	李鹏辉, 翟正利, 冯舒. 图对抗防御研究进展[J]. 计算机科学与探索, 2021, 15(12): 2292-2303.
[5]	蔺奇卡, 张玲玲, 刘均, 赵天哲. 基于问句感知图卷积的教育知识库问答方法[J]. 计算机科学与探索, 2021, 15(10): 1880-1887.