Collaborative recommendation algorithm based on deep graph neural network

doi:10.11772/j.issn.1001-9081.2022091361

Journal of Computer Applications ›› 2023, Vol. 43 ›› Issue (9): 2741-2746.DOI: 10.11772/j.issn.1001-9081.2022091361

• Artificial intelligence • Previous Articles Next Articles

Collaborative recommendation algorithm based on deep graph neural network

Runchao PAN, Qishan YU, Hongfei XIONG, Zhihui LIU()

School of Mathematics and Physics，China University of Geoscience，Wuhan Hubei 430074，China

Received:2022-09-12 Revised:2022-11-10 Accepted:2022-11-14 Online:2023-02-14 Published:2023-09-10
Contact: Zhihui LIU
About author:PAN Runchao， born in 1997， M. S. candidate. His research interests include recommendation system.
YU Qishan， born in 1997， M. S. candidate. His research interests include recommendation system.
XIONG Hongfei， born in 1997， M. S. candidate. His research interests include recommendation system.
Supported by:
National Natural Science Foundation of China(12001408);Science and Technology Research Project of Department of Education of Hubei Province(B2018541)

基于深度图神经网络的协同推荐算法

潘润超, 虞启山, 熊泓霏, 刘智慧()

中国地质大学（武汉）数学与物理学院，武汉 430074

通讯作者: 刘智慧
作者简介:潘润超（1997—），男，江苏常州人，硕士研究生，主要研究方向：推荐系统
虞启山（1997—），男，安徽马鞍山人，硕士研究生，主要研究方向：推荐系统
熊泓霏（1997—），男，湖北宜昌人，硕士研究生，主要研究方向：推荐系统；
基金资助:
国家自然科学基金资助项目(12001408);湖北省教育厅科学技术研究项目(B2018541)

Abstract

Abstract:

For the problem of over-smoothing in the existing recommendation algorithms based on Graph Neural Network （GNN）， a collaborative filtering recommendation algorithm based on deep GCN was proposed， namely Deep NGCF （Deep Neural Graph Collaborative Filtering）. In the algorithm， the initial residual connection and identity mapping were introduced into GNN， which avoided GNN from falling into over-smoothing after multiple graph convolution operations. Firstly， the initial embeddings of users and items were obtained through their interaction history. Next， in aggregation and propagation layer， collaborative signals of users and items in different stages were obtained with the use of initial residual connection and identity mapping. Finally， score prediction was performed according to the linear representation of all collaborative signals. In addition， to further improve the flexibility and recommendation performance of the model， the weights were set in the initial residual connection and identity mapping for adjustment. In order to verify the feasibility and effectiveness of Deep NGCF algorithm， experiments were conducted on datasets Gowalla， Yelp-2018 and Amazon-book. The results show that compared with the existing GNN recommendation algorithm such as Graph Convolutional Matrix Completion （GCMC） and Neural Graph Collaborate Filtering （NGCF）， Deep NGCF algorithm achieves the best results on recall and Normalized Discounted Cumulative Gain （NDCG）， thereby verifying the effectiveness of the proposed algorithm.

Key words: recommendation algorithm, collaborative filtering, Graph Neural Network (GNN), identity mapping, initial residual connection

摘要：

针对现有基于图神经网络（GNN）的推荐算法面临的过平滑的问题，提出一种基于深度GNN的协同过滤推荐算法Deep NGCF（Deep Neural Graph Collaborative Filtering）。该算法在GNN中引入初始残差连接和恒等映射，避免了GNN进行多次图卷积运算后陷入过平滑。首先，通过用户和项目的交互历史得到它们的初始嵌入；其次，在聚合传播层利用初始残差连接和恒等映射得到用户和项目的不同阶协同信号；最后，对所有协同信号进行线性表示以得到预测评分。此外，在初始残差连接和恒等映射中设置比重进行调节，从而进一步提高模型的灵活性和推荐性能。为验证Deep NGCF算法的可行性和有效性，在Gowalla、Yelp-2018与Amazon-Book数据集上进行实验。实验结果表明，相较于图卷积矩阵补全（GCMC）、神经图协同过滤（NGCF）等现有的GNN推荐算法，Deep NGCF算法取得了最高的召回率和归一化折损累计增益（NDCG），验证了所提算法的有效性。

CLC Number:

TP391

Runchao PAN, Qishan YU, Hongfei XIONG, Zhihui LIU. Collaborative recommendation algorithm based on deep graph neural network[J]. Journal of Computer Applications, 2023, 43(9): 2741-2746.

潘润超, 虞启山, 熊泓霏, 刘智慧. 基于深度图神经网络的协同推荐算法[J]. 《计算机应用》唯一官方网站, 2023, 43(9): 2741-2746.

Figures/Tables 9

Fig. 1 Bidirectional graph and its scoring matrix

Fig. 2 Flowchart of aggregation and propagation as well as score prediction

Tab. 1 Statistics of datasets

数据集	用户数	商品数	交互数	稀疏性/%
Gowalla	29 858	40 981	1 027 370	0.084
Yelp-2018	31 831	40 841	1 666 869	0.128
Amazon-Book	52 643	91 599	2 984 108	0.062

Tab. 2 Performance comparison of different recommendation algorithms

算法	Gowalla		Yelp-2018		Amazon-Book
算法	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$
BPRMF	0.129 1	0.110 9	0.043 3	0.035 4	0.025 0	0.019 6
GCMC	0.147 7	0.120 5	0.046 2	0.037 9	0.035 4	0.027 0
NGCF	0.153 0	0.131 3	0.057 9	0.047 7	0.034 4	0.025 3
LR-GCCF	0.151 8	0.125 9	0.056 1	0.034 3	0.034 1	0.025 8
LGC	0.179 0	0.151 8	0.063 9	0.052 5	0.039 6	0.029 8
Deep NGCF	0.180 7	0.152 1	0.064 5	0.053 2	0.041 7	0.033 2

Tab. 2 Performance comparison of different recommendation algorithms

算法	Gowalla		Yelp-2018		Amazon-Book
算法	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$
BPRMF	0.129 1	0.110 9	0.043 3	0.035 4	0.025 0	0.019 6
GCMC	0.147 7	0.120 5	0.046 2	0.037 9	0.035 4	0.027 0
NGCF	0.153 0	0.131 3	0.057 9	0.047 7	0.034 4	0.025 3
LR-GCCF	0.151 8	0.125 9	0.056 1	0.034 3	0.034 1	0.025 8
LGC	0.179 0	0.151 8	0.063 9	0.052 5	0.039 6	0.029 8
Deep NGCF	0.180 7	0.152 1	0.064 5	0.053 2	0.041 7	0.033 2

Tab. 3 Ablation experimental results

算法	Gowalla		Yelp-2018		Amazon-Book
算法	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$
A1	0.152 3	0.122 9	0.059 2	0.047 9	0.032 3	0.024 5
A2	0.150 5	0.121 3	0.058 3	0.047 2	0.029 8	0.022 9
A3	0.178 8	0.152 0	0.063 1	0.052 4	0.040 1	0.031 8
Deep NGCF	0.180 7	0.152 1	0.064 5	0.053 2	0.041 7	0.033 2

Tab. 3 Ablation experimental results

算法	Gowalla		Yelp-2018		Amazon-Book
算法	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$
A1	0.152 3	0.122 9	0.059 2	0.047 9	0.032 3	0.024 5
A2	0.150 5	0.121 3	0.058 3	0.047 2	0.029 8	0.022 9
A3	0.178 8	0.152 0	0.063 1	0.052 4	0.040 1	0.031 8
Deep NGCF	0.180 7	0.152 1	0.064 5	0.053 2	0.041 7	0.033 2

Tab. 4 Influence of number of network layers on model performance

网络层数	Gowalla		Yelp-2018		Amazon-Book
网络层数	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$
3	0.1781	0.1474	0.0598	0.0473	0.0384	0.0292
8	0.1767	0.1442	0.0607	0.0498	0.0395	0.0298
12	0.1792	0.1512	0.0622	0.0513	0.0416	0.0331
16	0.1807	0.1521	0.0645	0.0532	0.0417	0.0332
20	0.1785	0.1503	0.0633	0.0526	0.0402	0.0311

Tab. 4 Influence of number of network layers on model performance

网络层数	Gowalla		Yelp-2018		Amazon-Book
网络层数	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$	$R @ 20$	$N @ 20$
3	0.1781	0.1474	0.0598	0.0473	0.0384	0.0292
8	0.1767	0.1442	0.0607	0.0498	0.0395	0.0298
12	0.1792	0.1512	0.0622	0.0513	0.0416	0.0331
16	0.1807	0.1521	0.0645	0.0532	0.0417	0.0332
20	0.1785	0.1503	0.0633	0.0526	0.0402	0.0311

Fig. 3 Influence of regularization coefficient λ on performance

Fig. 4 Influence of hyperparameter α on performance

Fig. 5 Influence of hyperparameter β on performance

References 26

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.
	YU M， HE W T， ZHOU X C， et al. Review of recommendation system［J］. Journal of Computer Applications， 2022， 42（6）： 1898-1913.
2	宋瑞雪，李国勇. 基于改进的矩阵分解模型在推荐系统中的应用［J］. 计算机应用， 2019， 39（S1）： 93-95.
	SONG R X， LI G Y. Application of improved matrix factorization model in recommendation system［J］. Journal of Computer Applications， 2019， 39（S1）： 93-95.
3	SEDHAIN S， MENON A K， SANNER S， et al. AutoRec： autoencoders meet collaborative filtering［C］// Proceedings of the 24th International Conference on World Wide Web. Republic and Canton of Geneva： International World Wide Web Conferences Steering Committee， 2015： 111-112. 10.1145/2740908.2742726
4	CHENG H T， KOC L， HARMSEN J， et al. Wide & deep learning for recommender systems［C］// Proceedings of the 1st Workshop on Deep Learning for Recommender Systems. New York： ACM， 2016： 7-10. 10.1145/2988450.2988454
5	KIPF T N， WELLING M. Semi-supervised classification with graph convolutional networks［EB/OL］. （2017-02-22）［2022-03-02］.. 10.48550/arXiv.1609.02907
6	R van den BERG， KIPF T N， WELLING M. Graph convolutional matrix completion［C/OL］// Proceedings of the 2018 KDD Deep Learning Day ［2022-03-02］..
7	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 and Data Mining. New York： ACM， 2018：974-983. 10.1145/3219819.3219890
8	HAMILTON W L， YING R， LESKOVEC J. Inductive representation learning on large graphs［C］// Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook， NY： Curran Associates Inc.， 2017： 1025-1035. 10.7551/mitpress/11474.003.0014
9	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. 10.1145/3331184.3331267
10	WU F， ZHANG T Y， DE SOUZA A H. Jr.，et al. Simplifying graph convolutional networks［C］// Proceedings of the 36th International Conference on Machine Learning. New York： JMLR.org， 2019： 6861-6871.
11	CHEN L， WU L， HONG R C， et al. Revisiting graph based collaborative filtering： a linear residual graph convolutional network approach［C］// Proceedings of the 34th AAAI Conference on Artificial Intelligence. Palo Alto， CA： AAAI Press， 2020： 27-34. 10.1609/aaai.v34i01.5330
12	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. 10.1145/3397271.3401063
13	LI Q M， HAN Z C， WU X M. Deeper insights into graph convolutional networks for semi-supervised learning［C］// Proceedings of the 32nd AAAI Conference on Artificial Intelligence. Palo Alto， CA： AAAI Press， 2018： 3538-3545. 10.1609/aaai.v32i1.11604
14	LI G C H， MÜLLER M， THABET A， et al. DeepGCNs： can GCNs go as deep as CNNs？［C］// Proceedings of the 2019 IEEE/CVF International Conference on Computer Vision. Piscataway： IEEE， 2019： 9266-9275. 10.1109/iccv.2019.00936
15	CHIANG W L， LIU X Q， SI S， et al. Cluster-GCN： an efficient algorithm for training deep and large graph convolutional networks［C］// Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York： ACM， 2019： 257-266. 10.1145/3292500.3330925
16	CHEN M， WEI Z W， HUANG Z F， et al. Simple and deep graph convolutional networks［C］// Proceedings of the 37th International Conference on Machine Learning. New York： JMLR.org， 2020： 1725-1735.
17	张继杰，杨艳，刘勇. 利用初始残差和解耦操作的自适应深层图卷积［J］. 计算机应用， 2022， 42（1）：9-15. 10.11772/j.issn.1001-9081.2021071289
	ZHANG J J， YANG Y， LIU Y. Adaptive deep graph convolution using initial residual and decoupling operations［J］. Journal of Computer Applications， 2022， 42（1）： 9-15. 10.11772/j.issn.1001-9081.2021071289
18	杨壮，吴斌，廉炜雯，等. 基于注意力机制和深度恒等映射的人脸识别［J］. 传感器与微系统， 2020， 39（9）：150-153.
	YANG Z， WU B， LIAN W W， et al. Face recognition based on attention mechanism and deep equivariant mapping［J］. Transducer and Microsystem Technologies， 2020， 39（9）： 150-153.
19	OONO K， SUZUKI T. Graph neural networks exponentially lose expressive power for node classification［EB/OL］. （2021-01-06）［2022-03-02］..
20	RENDLE S， FREUDENTHALER C， GANTER Z， et al. BPR： Bayesian personalized ranking from implicit feedback［C］// Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence. Arlington， VA： AUAI Press， 2009： 452-461.
21	LIANG D W， CHARLIN L， McINERNEY J， et al. Modeling user exposure in recommendation［C］// Proceedings of the 25th International Conference on World Wide Web. Republic and Canton of Geneva： International World Wide Web Conferences Steering Committee， 2016： 951-961. 10.1145/2872427.2883090
22	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. Republic and Canton of Geneva： International World Wide Web Conferences Steering Committee， 2017： 173-182. 10.1145/3038912.3052569
23	HE R N， McAULEY J. Ups and downs： modeling the visual evolution of fashion trends with one-class collaborative filtering［C］// Proceedings of the 25th International Conference on World Wide Web. Republic and Canton of Geneva： International World Wide Web Conferences Steering Committee， 2016： 507-517. 10.1145/2872427.2883037
24	赵若宇. 图神经网络在推荐系统中的应用［D］. 合肥：中国科学技术大学， 2021：1-49. 10.47056/0365-9615-2021-172-12-719-723
	ZHAO R Y. Application of graph neural networks in recommendation systems［D］. Hefei： University of Science and Technology of China， 2021：1-49. 10.47056/0365-9615-2021-172-12-719-723
25	张晟铭. 基于图神经网络的推荐系统研究［D］. 成都：电子科技大学， 2021：1-65.
	ZHANG S M. Research on recommender systems using graph neural networks［D］. Chengdu： University of Electronic Science and Technology of China， 2021：1-65.
26	肖海涛，何鹏，曾诚. 基于排序学习的Top-k软件服务推荐方法［J］. 计算机应用， 2018， 38（S1）： 144-149， 169.
	XIAO H T， HE P， ZENG C. Top-k software service recommendation method based on learning to rank［J］. Journal of Computer Applications， 2018， 38（S1）： 144-149， 169.

[1]	Haodong ZHENG, Hua MA, Yingchao XIE, Wensheng TANG. Knowledge tracing model based on graph neural network blending with forgetting factors and memory gate [J]. Journal of Computer Applications, 2023, 43(9): 2747-2752.
[2]	Hongjun HENG, Dingcheng YANG. Knowledge enhanced aspect word interactive graph neural network [J]. Journal of Computer Applications, 2023, 43(8): 2412-2419.
[3]	Kun ZHANG, Fengyu YANG, Fa ZHONG, Guangdong ZENG, Shijian ZHOU. Source code vulnerability detection based on hybrid code representation [J]. Journal of Computer Applications, 2023, 43(8): 2517-2526.
[4]	Zifang XIA, Yaxin YU, Ziteng WANG, Jiaqi QIAO. Explainable recommendation mechanism by fusion collaborative knowledge graph and counterfactual inference [J]. Journal of Computer Applications, 2023, 43(7): 2001-2009.
[5]	Kejun JIN, Hongtao YU, Yiteng WU, Shaomei LI, Jianpeng ZHANG, Honghao ZHENG. Improved defense method for graph convolutional network based on singular value decomposition [J]. Journal of Computer Applications, 2023, 43(5): 1511-1517.
[6]	Hao SUN, Jian CAO, Haisheng LI, Dianhui MAO. Session-based recommendation model based on enhanced capsule network [J]. Journal of Computer Applications, 2023, 43(4): 1043-1049.
[7]	Lubao LI, Tian CHEN, Fuji REN, Beibei LUO. Bimodal emotion recognition method based on graph neural network and attention [J]. Journal of Computer Applications, 2023, 43(3): 700-705.
[8]	Yu WANG, Yubo YUAN, Yi GUO, Jiajie ZHANG. Sentiment boosting model for emotion recognition in conversation text [J]. Journal of Computer Applications, 2023, 43(3): 706-712.
[9]	ZHANG Jun, WU Pengli, SHI Lukui, SHI Jin, PAN Bin. Deep learning model for multi-station temperature prediction combined with MOD11A1 and surface meteorological station data [J]. Journal of Computer Applications, 2023, 43(1): 321-328.
[10]	Meng YU, Wentao HE, Xuchuan ZHOU, Mengtian CUI, Keqi WU, Wenjie ZHOU. Review of recommendation system [J]. Journal of Computer Applications, 2022, 42(6): 1898-1913.
[11]	Jiafan ZHOU, Yuefeng DU, Baoyan SONG, Xiaoguang LI, Azhu ZHAO, Xujie XIAO. MOOC video recommendation method based on meta-path attention mechanism [J]. Journal of Computer Applications, 2022, 42(6): 1808-1813.
[12]	Xiaohan LI, Huading JIA, Xue CHENG, Taiyong LI. Stock market volatility prediction method based on improved genetic algorithm and graph neural network [J]. Journal of Computer Applications, 2022, 42(5): 1624-1633.
[13]	Suqi ZHANG, Xinxin WANG, Shiyao SHE, Junhua GU. Knowledge graph recommendation model with multiple time scales and feature enhancement [J]. Journal of Computer Applications, 2022, 42(4): 1093-1098.
[14]	Junhua GU, Rui WANG, Ningning LI, Suqi ZHANG. Knowledge graph attention network fusing collaborative filtering information [J]. Journal of Computer Applications, 2022, 42(4): 1087-1092.
[15]	Wen WEN, Fangyu LIANG. Sequential behavior recommendation based on user’s latent state and dependency learning [J]. Journal of Computer Applications, 2022, 42(12): 3756-3762.

Collaborative recommendation algorithm based on deep graph neural network

基于深度图神经网络的协同推荐算法

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 9

References 26

Related Articles 15

Recommended Articles

Metrics