基于动态邻域采样的社交推荐模型

doi:10.12141/j.issn.1000-565X.230023

摘要/Abstract

摘要：

基于图神经网络的社交推荐模型在提升推荐系统性能方面有不错的表现。但现有方法都忽略了被查询的目标用户和项目节点与其邻居节点间可能存在特征不匹配的问题，导致噪声的引入而降低了模型性能。为了解决该问题，文中提出一种社交推荐模型DNSSR。首先构建一个包含用户和项目多元关系的关系图谱，图节点间信息关联更丰富；然后利用动态邻域采样机制获得与目标查询对的特征更一致的邻居节点，减少了噪声信息；另外，为了进一步提高模型预测性能，设计了一种增强型图神经网络对采样后得到的关系子图进行建模，它可以区分不同邻居节点的重要性并选择更可靠的信息源，获得更鲁棒的用户和项目嵌入向量用于评分预测。实验结果表明：相比其他先进模型，该模型预测误差明显降低，证明了文中所提各项方法的有效性；尤其是动态邻域采样机制，若将其弃用，DNSSR在Ciao数据集上的RMSE（均方根误差）和MAE（平均绝对误差）指标将分别上升6.05%和7.31%，在Epinions数据集上则分别上升3.49%和5.41%，充分验证了其能有效降低噪声干扰、提高社交推荐模型的性能。

关键词: 社交推荐, 评分预测, 图神经网络, 动态邻域采样

Abstract:

The social recommendation model based on graph neural network has achieved good performance in improving the performance of the recommendation system. However, the existing methods ignored the possible feature mismatch between the queried target users and content nodes and their neighbors, which leads to the introduction of noise and reduces the model performance. To solve this problem, this paper proposed a social recommendation model DNSSR. Firstly, it constructed a relational graph containing multiple relationships between users and items, with richer information associations between nodes in the graph. Then the dynamic neighborhood sampling mechanism was used to obtain neighbor nodes that are more consistent with the characteristics of the target query pair, reducing noise information. In addition, in order to further improve the predictive performance of the model, this paper designed an enhanced graph neural network to model the sampled relationship subgraphs. It can distinguish the importance of different neighboring nodes and select more reliable information sources to obtain more robust user and item embedding vectors for rating prediction. The experimental results show that the prediction error of this model is significantly reduced compared to that of other advanced models, proving the effectiveness of the methods proposed in the paper. Especially for the dynamic neighborhood sampling mechanism, if it is abandoned, the RMSE and MAE indicators of DNSSR on the Ciao dataset will increase by 6.05% and 7.31% respectively, and the Epinions dataset will increase by 3.49% and 5.41% respectively, which fully demonstrate their effectiveness in reducing noise interference and improving the performance of social recommendation models.

Key words: social recommendation, rating prediction, graph neural network, dynamic neighborhood sampling

中图分类号:

TP391

蔡晓东, 周青松, 叶青. 基于动态邻域采样的社交推荐模型[J]. 华南理工大学学报(自然科学版), 2024, 52(2): 32-41.

CAI Xiaodong, ZHOU Qingsong, YE Qing. Social Recommendation Model Based on Dynamic Neighborhood Sampling[J]. Journal of South China University of Technology(Natural Science Edition), 2024, 52(2): 32-41.

图/表 8

图1

图2

图3

表1

表2

图4

表3

图5

参考文献 24

1	DA’U A， SALIM N ．Recommendation system based on deep learning methods：asystematic review and new directions［J］．Artificial Intelligence Review，2020，53（4）：2709-2748.
2	JANNACH D， ZANKER M， FELFERNIG A，et al ．Recommender systems：an introduction［M］．Cambridge：Cambridge University Press，2010.
3	MCPHERSON M， SMITH-LOVIN L， COOK J M ．Birds of a feather：homophily in social networks［J］．Annual Review of Sociology，2001，27（1）：415-444.
4	SALAKHUTDINOV R， MNIH A ．Probabilistic matrix factorization［C］∥Proceedings of the 20th International Conference on Neural Information Processing Systems．Vancouver：Curran Associates Inc.，2007：1257-1264.
5	MA H， YANG H， LYU M R，et al ．Sorec：social recommendation using probabilistic matrix factorization［C］∥Proceedings of the 17th ACM Conference on Information and Knowledge Management．California：ACM，2008：931-940.
6	JAMALI M， ESTER M ．A matrix factorization technique with trust propagation for recommendation in social networks［C］∥Proceedings of the Fourth ACM Conference on Recommender Systems．New York：ACM，2010：135-142.
7	YANG B， LEI Y， LIU J，et al ．Social collaborative filtering by trust［J］．IEEE Transactions on Pattern Analysis and Machine Intelligence，2016，39（8）：1633-1647.
8	KIPF T N， WELLING M ．Semi-supervised classification with graph convolutional networks［C］∥Proceedings of the 5th International Conference on Learning Representations．Toulon：ICLR，2017：1-14.
9	WU Z， PAN S， CHEN F，et al ．A comprehensive survey on graph neural networks［J］．IEEE Transactions on Neural Networks and Learning Systems，2020，32（1）：4-24.
10	ZHOU J， CUI G， HU S，et al ．Graph neural networks：a review of methods and applications［J］．AI Open，2020，1：57-81.
11	FAN W， MA Y， LI Q，et al ．Graph neural networks for social recommendation［C］∥Proceedings of the World Wide Web Conference．San Francisco：ACM，2019：417-426.
12	WU Q， ZHANG H， GAO X，et al ．Dual graph attention networks for deep latent representation of multifaceted social effects in recommender systems［C］∥Proceedings of the World Wide Web Conference．San Francisco：ACM，2019：2091-2102.
13	XIAO Y， YAO L， PEI Q，et al ．MGNN：mutualistic graph neural network for joint friend and item recommendation［J］．IEEE Intelligent Systems，2020，35（5）：7-17.
14	FAN W， MA Y， LI Q，et al ．A graph neural network framework for social recommendations［J］．IEEE Transactions on Knowledge and Data Engineering，2022，34（5）：2033-2047.
15	ZHANG Y， WU L， SHEN Q，et al ．Graph learning augmented heterogeneous graph neural network for social recommendation［J］．ACM Transactions on Recommender Systems，2023，1（4）：1-22.
16	YU Y H， QIAN W W， ZHANG L，et al ．A graph-neural-network-based social network recommendation algorithm using high-order neighbor information［J］．Sensors，2022，22（19）：7122/1-16.
17	CHEN J， XIN X， LIANG X，et al ．GDSRec：graph-based decentralized collaborative filtering for social recommendation［J］．IEEE Transactions on Knowledge and Data Engineering，2023，35（5）：4813-4824.
18	HUANG C， XU H， XU Y，et al ．Knowledge-aware coupled graph neural network for social recommendation［J］．Proceedings of the AAAI Conference on Artificial Intelligence，2021，35（5）：4115-4122.
19	FU B， ZHANG W， HU G，et al ．Dual side deep context-aware modulation for social recommendation［C］∥Proceedings of the Web Conference 2021．Ljubljana：Association for Computing Machinery，2021：2524-2534.
20	YANG L， LIU Z， DOU Y，et al ．Consisrec：enhancing GNN for social recommendation via consistent neighbor aggregation［C］∥Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval．Montréal：ACM，2021：2141-2145.
21	CHEN T， WONG R C W ．An efficient and effective framework for session-based social recommendation［C］∥Proceedings of the 14th ACM International Conference on Web Search and Data Mining．Israel：ACM，2021：400-408.
22	WANG Z， WEI W， CONG G，et al ．Global context enhanced graph neural networks for session-based recommendation［C］∥Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval．Xi’an：ACM，2020：169-178.
23	FAN W， LI Q， CHENG M ．Deep modeling of social relations for recommendation［C］∥Proceedings of the AAAI Conference on Artificial Intelligence．New Orleans：AAAI，2018，32：8075-8076.
24	KOREN Y ．Factorization meets the neighborhood：amultifaceted collaborative filtering model［C］∥Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining．New York：ACM，2008：426-434.

统计信息	数值
统计信息	Ciao	Epinions
用户数量	7 317	18 088
项目数量	104 975	261 649
评分数量	283 319	764 352
评分稠密度/%	0.036 8	0.016 1
社交关系数量	111 781	355 813
社交关系稠密度/%	0.208 7	0.108 7
项目类别数量	28	27

模型	Ciao		Epinions
模型	RMSE	MAE	RMSE	MAE
PMF	1.123 8	0.902 1	1.212 8	0.995 2
SoRec	1.065 2	0.841 0	1.143 7	0.896 1
TrustMF	1.047 9	0.769 0	1.139 5	0.841 0
DeepSoR	1.031 6	0.773 9	1.097 2	0.838 3
GraphRec	0.979 4	0.738 7	1.063 1	0.816 8
GraphRec+	0.978 7	0.733 0	1.057 6	0.809 3
GDSRec	0.974 0	0.732 3	1.056 6	0.804 7
DNSSR	0.951 3	0.714 4	1.045 9	0.794 4

模型	Ciao		Epinions
模型	RMSE	MAE	RMSE	MAE
DNSSR	0.951 3	0.714 4	1.045 9	0.794 4
DNSSR-nan	0.966 2	0.734 1	1.061 5	0.814 7
DNSSR-cat	0.959 8	0.725 9	1.052 4	0.806 6
DNSSR-add	0.963 1	0.731 2	1.057 8	0.813 2

[1]	赵荣超, 吴百礼, 陈祝云, 温楷儒, 张绍辉, 李巍华. 多尺度时空信息融合驱动的图神经网络故障诊断方法[J]. 华南理工大学学报(自然科学版), 2023, 51(12): 42-52.
[2]	许伦辉, 余佳芯, 裴明阳, 等. 基于几何路网结构和强化学习的车辆重定位策略[J]. 华南理工大学学报(自然科学版), 2023, 51(10): 99-109.
[3]	陆以勤, 潘周双, 张洋, 等. 基于交互式连接图注意力网络的知识图谱补全方法[J]. 华南理工大学学报(自然科学版), 2022, 50(12): 13-19.
[4]	蔡晓东, 曾志杨. AFGSRec：一种自适应融合全局协同特征的社交推荐模型[J]. 华南理工大学学报(自然科学版), 2022, 50(12): 71-79.
[5]	古万荣, 谢贤芬, 张子烨, 等. 面向非随机缺失数据的协同过滤评分方法[J]. 华南理工大学学报（自然科学版）, 2021, 49(1): 47-57.