Recommendation algorithm of graph contrastive learning based on hybrid negative sampling

doi:10.11772/j.issn.1001-9081.2024040419

Journal of Computer Applications ›› 2025, Vol. 45 ›› Issue (4): 1053-1060.DOI: 10.11772/j.issn.1001-9081.2024040419

• Artificial intelligence • Previous Articles Next Articles

Recommendation algorithm of graph contrastive learning based on hybrid negative sampling

Renjie TIAN¹, Mingli JING¹(), Long JIAO², Fei WANG¹

^1.School of Electronic Engineering，Xi’an Shiyou University，Xi’an Shaanxi 710065，China
^2.College of Chemical Engineering，Xi’an Shiyou University，Xi’an Shaanxi 710065，China

Received:2024-04-10 Revised:2024-07-15 Accepted:2024-07-18 Online:2025-04-08 Published:2025-04-10
Contact: Mingli JING
About author:TIAN Renjie， born in 2000， M. S. candidate. His research interests include recommendation system.
JIAO Long， born in 1980， Ph. D.， professor. His research interests include chemometrics， pattern recognition.
WANG Fei， born in 1985， Ph. D.， lecturer. His research interests include image processing， signal processing， algorithm optimization.
First author contact:JING Mingli， born in 1977， Ph. D.， associate professor. His research interests include sparse low-rank recovery in compressive sensing， numerical optimization algorithm， image processing， machine learning.
Supported by:
National Natural Science Foundation of China(22373075);Key Research and Development Program of Shaanxi Province(2022GY-435);Xi’an Shiyou University Graduate Innovation and Practical Ability Cultivation Program(YCS23114145)

基于混合负采样的图对比学习推荐算法

田仁杰¹, 景明利¹(), 焦龙², 王飞¹

^1.西安石油大学电子工程学院，西安 710065
^2.西安石油大学化工学院，西安 710065

通讯作者: 景明利
作者简介:田仁杰（2000—），男，陕西渭南人，硕士研究生，主要研究方向：推荐系统
焦龙（1980—），男，陕西西安人，教授，博士，主要研究方向：化学计量学、模式识别
王飞（1985—），男，陕西宝鸡人，讲师，博士，主要研究方向：图像处理、信号处理、算法优化。
第一联系人：景明利（1977—），男，陕西长武人，副教授，博士，主要研究方向：压缩感知中的稀疏低秩恢复、数值优化算法、图像处理、机器学习
基金资助:
国家自然科学基金资助项目(22373075);陕西省重点研发计划项目(2022GY-435);西安石油大学研究生创新与实践能力培养计划项目(YCS23114145)

Abstract

Abstract:

Contrastive Learning （CL） has the ability to extract self-supervised signals from raw data， providing strong support for addressing data sparsity in recommender systems. However， most existing CL-based recommendation algorithms focus on improving model structures and data augmentation methods， and ignoring the importance of enhancing negative sample quality and uncovering potential implicit relationships between users and items in recommendation tasks. To address this issue， a Hybrid negative Sampling-based Graph Contrastive Learning recommendation algorithm （HSGCL） was proposed. Firstly， differing from the uniform sampling method to sample from real data， a positive sample hybrid method was used by the proposed algorithm to integrate positive sample information into negative samples. Secondly， informative hard negative samples were created through a skip-mix method. Meanwhile， multiple views were generated by altering the graph structure using Node Dropout （ND）， and controlled uniform noise smoothing was introduced in the embedding space to adjust the uniformity of learning representations. Finally， the main recommendation task and CL task were trained jointly. Numerical experiments were conducted on three public datasets： Douban-Book， Yelp2018， and Amazon-Kindle. The results show that compared to the baseline model Light Graph Convolution Network （LightGCN）， the proposed algorithm improves the Recall@20 by 23%， 13%， and 7%， respectively， and Normalized Discounted Cumulative Gain （NDCG@20） by 32%， 14%， and 5%， respectively， and performs excellently in enhancing the diversity of negative sample embedding information. It can be seen that by improving negative sampling method and data augmentation， the proposed algorithm improves negative sample quality， the uniformity of representation distribution， and accuracy of the recommendation algorithm.

Key words: Graph Neural Network (GNN), Contrastive Learning (CL), recommender system, negative sampling, data augmentation

摘要：

对比学习（CL）具有可从原始数据中提取自监督信号的特性，为推荐系统解决数据稀疏问题提供了有力支持。然而，现有的CL推荐算法大多着眼于改进模型结构和数据增强方法，忽视了提升推荐任务中的负样本质量以及挖掘用户与项目之间潜在隐性关系的重要性。针对此问题，提出一种基于混合负采样的图对比学习推荐算法（HSGCL）。首先，与均匀采样方法从真实数据中采样不同，所提算法使用正样本混合方法将正样本信息融入负样本中；其次，通过跳跃混合方法创造富含信息的难负样本；同时，通过使用节点丢弃（ND），改变图结构以生成多个视图，并在嵌入空间中引入可控的均匀噪声平滑调整学习表示的均匀性；最后，将推荐主任务与CL任务进行联合训练。在Douban-Book、Yelp2018和Amazon-Kindle这3个公共数据集上的数值实验结果表明，相较于基线模型——轻量化图卷积网络（LightGCN），所提算法在召回率（Recall@20）上分别提升了23%、13%和7%，在归一化折损累积增益（NDCG@20）上分别提升了32%、14%和5%，且在提升负样本嵌入信息多样性方面表现优异。可见，所提算法从负采样方法和数据增强两方面进行改进，提高了负样本质量、表示分布的均匀性和推荐算法的准确性。

关键词: 图神经网络, 对比学习, 推荐系统, 负采样, 数据增强

CLC Number:

TP391.3

Renjie TIAN, Mingli JING, Long JIAO, Fei WANG. Recommendation algorithm of graph contrastive learning based on hybrid negative sampling[J]. Journal of Computer Applications, 2025, 45(4): 1053-1060.

田仁杰, 景明利, 焦龙, 王飞. 基于混合负采样的图对比学习推荐算法[J]. 《计算机应用》唯一官方网站, 2025, 45(4): 1053-1060.

Figures/Tables 9

References 34

1	ZHANG S， TONG H， XU J， et al. Graph convolutional networks： a comprehensive review［J］. Computational Social Networks， 2019， 6： No.11.
2	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.
3	孙轩宇，史艳翠. 融合项目影响力的图神经网络会话推荐模型［J］. 计算机应用， 2023， 43（12）： 3689-3696.
	SUN X Y， SHI Y C. Session-based recommendation model by graph neural network fused with item influence［J］. Journal of Computer Applications， 2023， 43（12）： 3689-3696.
4	潘润超，虞启山，熊泓霏，等. 基于深度图神经网络的协同推荐算法［J］. 计算机应用， 2023， 43（9）： 2741-2746.
	PAN R C， YU Q S， XIONG H F， et al. Collaborative recommendation algorithm based on deep graph neural network［J］. Journal of Computer Applications， 2023， 43（9）： 2741-2746.
5	WANG X， HE X， 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.
6	HE X， 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.
7	LIU Z， MA Y， OUYANG Y， et al. Contrastive learning for recommender system［EB/OL］. ［2024-04-01］..
8	YIN Y， WANG Q， HUANG S， et al. AutoGCL： automated graph contrastive learning via learnable view generators［C］// Proceedings of the 36th AAAI Conference on Artificial Intelligence. Palo Alto： AAAI Press， 2022： 8892-8900.
9	韩滕跃，牛少彰，张文. 基于对比学习的多模态序列推荐算法［J］. 计算机应用， 2022， 42（6）： 1683-1688.
	HAN T Y， NIU S Z， ZHANG W. Multimodal sequential recommendation algorithm based on contrastive learning［J］. Journal of Computer Applications， 2022， 42（6）： 1683-1688.
10	SHEN X， SUN D， PAN S， et al. Neighbor contrastive learning on learnable graph augmentation［C］// Proceedings of the 37th AAAI Conference on Artificial Intelligence. Palo Alto： AAAI Press， 2023： 9782-9791.
11	ZHAO Q， WU Z， ZHANG Z， et al. Long-tail augmented graph contrastive learning for recommendation［C］// Proceedings of the 2023 Joint European Conference on Machine Learning and Knowledge Discovery in Databases， LNCS 14172. Cham： Springer， 2023： 387-403.
12	LIANG K， LIU Y， ZHOU S， et al. Knowledge graph contrastive learning based on relation-symmetrical structure［J］. IEEE Transactions on Knowledge and Data Engineering， 2023， 36（1）： 226-238.
13	CHEN M， HUANG C， XIA L， et al. Heterogeneous graph contrastive learning for recommendation［C］// Proceedings of the 16th ACM International Conference on Web Search and Data Mining. New York： ACM， 2023： 544-552.
14	WU J， WANG X， FENG F， et al. Self-supervised graph learning for recommendation［C］// Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York： ACM， 2021： 726-735.
15	CAI X， HUANG C， XIA L， et al. LightGCL： simple yet effective graph contrastive learning for recommendation［EB/OL］. ［2024-04-01］..
16	YU J， YIN H， XIA X， et al. Are graph augmentations necessary？ simple graph contrastive learning for recommendation［C］// Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York： ACM， 2022： 1294-1303.
17	YU J， XIA X， CHEN T， et al. XSimGCL： towards extremely simple graph contrastive learning for recommendation［J］. IEEE Transactions on Knowledge and Data Engineering， 2024， 36（2）： 913-926.
18	HUANG T， DONG Y， DING M， et al. MixGCF： an improved training method for graph neural network-based recommender systems［C］// Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York： ACM， 2021： 665-674.
19	LIN Z， TIAN C， HOU Y， et al. Improving graph collaborative filtering with neighborhood-enriched contrastive learning［C］// Proceedings of the ACM Web Conference 2022. New York： ACM， 2022： 2320-2329.
20	JIANG Y， HUANG C， HUANG L. Adaptive graph contrastive learning for recommendation［C］// Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York： ACM， 2023： 4252-4261.
21	ZHOU K， WANG H， ZHAO W X， et al. S³-Rec： self-supervised learning for sequential recommendation with mutual information maximization［C］// Proceedings of the 29th ACM International Conference on Information and Knowledge Management. New York： ACM， 2020： 1893-1902.
22	WEI Y， WANG X， LI Q， et al. Contrastive learning for cold-start recommendation［C］// Proceedings of the 29th ACM International Conference on Multimedia. New York： ACM， 2021： 5382-5390.
23	QIU R， HUANG Z， YIN H， et al. Contrastive learning for representation degeneration problem in sequential recommendation［C］// Proceedings of the 15th ACM International Conference on Web Search and Data Mining. New York： ACM， 2022： 813-823.
24	YOU Y， CHEN T， SUI Y， et al. Graph contrastive learning with augmentations［C］// Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook： Curran Associates Inc.， 2020： 5812-5823.
25	范伟，周魏，文俊浩. 基于异构图的双通道交叉自适应对比学习推荐［J］. 电子学报， 2023， 51（7）：1929-1938.
	FAN W， ZHOU W， WEN J H. Recommendation based on heterogeneous graph using dual channel cross-adaptive contrastive learning［J］. Acta Electronica Sinica， 2023， 51（7）： 1929-1938.
26	YANG Z， DING M， ZHOU C， et al. Understanding negative sampling in graph representation learning［C］// Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York： ACM， 2020： 1666-1676.
27	黄河源，慕彩红，方云飞，等. 使用图负采样的图卷积神经网络推荐算法［J］. 西安电子科技大学学报， 2024， 51（1）：86-99.
	HUANG H Y， MU C H， FANG Y F， et al. Graph convolution neural network for recommendation using graph negative sampling［J］. Journal of Xidian University， 2024， 51（1）：86-99.
28	KALANTIDIS Y， SARIYILDIZ M B， PION N， et al. Hard negative mixing for contrastive learning［C］// Proceedings of the 34th International Conference on Neural Information Processing Systems. Red Hook： Curran Associates Inc.， 2020： 21798-21809.
29	ZHANG H， CISSE M， DAUPHIN Y N， et al. mixup： Beyond empirical risk minimization［EB/OL］. ［2024-04-01］..
30	ZHANG W， CHEN T， WANG J， et al. Optimizing top-N collaborative filtering via dynamic negative item sampling［C］// Proceedings of the 36th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York： ACM， 2013： 785-788.
31	GUTMANN M， HYVÄRINEN A. Noise-contrastive estimation： a new estimation principle for unnormalized statistical models［C］// Proceedings of the 13th International Conference on Artificial Intelligence and Statistics. New York： JMLR.org， 2010： 297-304.
32	YAO T， YI X， CHENG D Z， et al. Self-supervised learning for large-scale item recommendations［C］// Proceedings of the 30th ACM International Conference on Information and Knowledge Management. New York： ACM， 2021： 4321-4330.
33	LEE D， KANG S， JU H， et al. Bootstrapping user and item representations for one-class collaborative filtering［C］// Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York： ACM， 2021： 317-326.
34	袁琮淇，刘渊，刘静文. 基于混合采样的图对比学习推荐算法［J］. 计算机应用研究， 2023， 40（5）： 1346-1351.
	YUAN C Q， LIU Y， LIU J W. Graph contrastive learning recommendation algorithm based on mixed sampling［J］. Application Research of Computers， 2023， 40（5）： 1346-1351.

数据集	用户数	项目数	交互数	密集度/%
Douban-Book	13 024	22 347	792 062	0.272
Yelp2018	31 668	38 048	1 561 406	0.130
Amazon-Kindle	138 333	98 572	1 909 965	0.014

数据集	用户数	项目数	交互数	密集度/%
Douban-Book	13 024	22 347	792 062	0.272
Yelp2018	31 668	38 048	1 561 406	0.130
Amazon-Kindle	138 333	98 572	1 909 965	0.014

算法	Douban-Book		Yelp2018		Amazon-Kindle
算法	R@20	N@20	R@20	N@20	R@20	N@20
LightGCN	0.150 1	0.128 2	0.063 9	0.052 5	0.205 7	0.131 5
SGL	0.173 2	0.155 1	0.067 5	0.055 5	0.210 5	0.135 1
DNN+SSL	0.136 6	0.114 8	0.048 3	0.038 2	0.152 0	0.098 9
BUIR	0.112 7	0.089 3	0.048 7	0.040 4	0.092 2	0.052 8
MSGCL	0.176 3	0.158 0	0.068 8	0.055 8	0.210 1	0.128 3
MixGCF	0.173 1	0.155 2	0.071 3	0.058 4	0.212 8	0.132 7
SimGCL	0.177 1	0.158 1	0.071 9	0.059 7	0.210 2	0.136 5
AdaGCL	0.183 1	0.166 6	0.071 5	0.059 4	0.215 9	0.136 7
HSGCL	0.184 1 （+23%）	0.169 3 （+32%）	0.071 9 （+13%）	0.059 9 （+14%）	0.219 5 （+7%）	0.137 6 （+5%）

算法	Douban-Book		Yelp2018		Amazon-Kindle
算法	R@20	N@20	R@20	N@20	R@20	N@20
LightGCN	0.150 1	0.128 2	0.063 9	0.052 5	0.205 7	0.131 5
SGL	0.173 2	0.155 1	0.067 5	0.055 5	0.210 5	0.135 1
DNN+SSL	0.136 6	0.114 8	0.048 3	0.038 2	0.152 0	0.098 9
BUIR	0.112 7	0.089 3	0.048 7	0.040 4	0.092 2	0.052 8
MSGCL	0.176 3	0.158 0	0.068 8	0.055 8	0.210 1	0.128 3
MixGCF	0.173 1	0.155 2	0.071 3	0.058 4	0.212 8	0.132 7
SimGCL	0.177 1	0.158 1	0.071 9	0.059 7	0.210 2	0.136 5
AdaGCL	0.183 1	0.166 6	0.071 5	0.059 4	0.215 9	0.136 7
HSGCL	0.184 1 （+23%）	0.169 3 （+32%）	0.071 9 （+13%）	0.059 9 （+14%）	0.219 5 （+7%）	0.137 6 （+5%）

变体	Douban-Book		Yelp2018		Amazon-Kindle
变体	R@20	N@20	R@20	N@20	R@20	N@20
HSGCL_a	0.162 2	0.147 5	0.059 8	0.049 5	0.129 2	0.081 0
HSGCL_p	0.182 1	0.167 4	0.070 2	0.058 2	0.216 3	0.135 9
HSGCL_g	0.183 9	0.169 0	0.071 2	0.058 7	0.218 2	0.137 1

Recommendation algorithm of graph contrastive learning based on hybrid negative sampling

基于混合负采样的图对比学习推荐算法

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 9

References 34

Related Articles 15

Recommended Articles

Metrics

[1]	Cong WANG, Yancui SHI. Group recommendation model by graph neural network based on multi-perspective learning [J]. Journal of Computer Applications, 2025, 45(4): 1205-1212.
[2]	Lan YOU, Yuang ZHANG, Yuan LIU, Zhijun CHEN, Wei WANG, Xing ZENG, Zhangwei HE. Developer recommendation for open-source projects based on collaborative contribution network [J]. Journal of Computer Applications, 2025, 45(4): 1213-1222.
[3]	Haitao SUN, Jiayu LIN, Zuhong LIANG, Jie GUO. Data augmentation technique incorporating label confusion for Chinese text classification [J]. Journal of Computer Applications, 2025, 45(4): 1113-1119.
[4]	Weichao DANG, Xinyu WEN, Gaimei GAO, Chunxia LIU. Multi-view and multi-scale contrastive learning for graph collaborative filtering [J]. Journal of Computer Applications, 2025, 45(4): 1061-1068.
[5]	Chenwei SUN, Junli HOU, Xianggen LIU, Jiancheng LYU. Large language model prompt generation method for engineering drawing understanding [J]. Journal of Computer Applications, 2025, 45(3): 801-807.
[6]	Kun FU, Shicong YING, Tingting ZHENG, Jiajie QU, Jingyuan CUI, Jianwei LI. Graph data augmentation method for few-shot node classification [J]. Journal of Computer Applications, 2025, 45(2): 392-402.
[7]	Handa MA, Yadong WU. Multi-domain spatiotemporal hierarchical graph neural network for air quality prediction [J]. Journal of Computer Applications, 2025, 45(2): 444-452.
[8]	Qijian CAI, Wei TAN. Semantic graph enhanced multi-modal recommendation algorithm [J]. Journal of Computer Applications, 2025, 45(2): 421-427.
[9]	Xuewen YAN, Zhangjin HUANG. Few-shot image classification method based on contrast learning [J]. Journal of Computer Applications, 2025, 45(2): 383-391.
[10]	Zidong CHENG, Peng LI, Feng ZHU. Potential relation mining in internet of things threat intelligence knowledge graph [J]. Journal of Computer Applications, 2025, 45(1): 24-31.
[11]	Wenbo ZHAO, Zitong MA, Zhe YANG. Link prediction model based on directed hypergraph adaptive convolution [J]. Journal of Computer Applications, 2025, 45(1): 15-23.
[12]	Jialin ZHANG, Qinghua REN, Qirong MAO. Speaker verification system utilizing global-local feature dependency for anti-spoofing [J]. Journal of Computer Applications, 2025, 45(1): 308-317.
[13]	Tingjie TANG, Jiajin HUANG, Jin QIN. Session-based recommendation with graph auxiliary learning [J]. Journal of Computer Applications, 2024, 44(9): 2711-2718.
[14]	Xinrui LIN, Xiaofei WANG, Yan ZHU. Academic anomaly citation group detection based on local extended community detection [J]. Journal of Computer Applications, 2024, 44(6): 1855-1861.
[15]	Jiong WANG, Taotao TANG, Caiyan JIA. PAGCL： positive augmentation graph contrastive learning recommendation method without negative sampling [J]. Journal of Computer Applications, 2024, 44(5): 1485-1492.