Collaborative Filtering Recommendation Model Based on Improved Graph Convolutional Network

doi:10.3778/j.issn.1002-8331.2101-0324

Abstract

Abstract: graph convolutional networks（GCN） have been applied in various fields with great success, including recommendation systems. The core goal of the recommendation system is to infer user preferences so as to actively provide users with valuable and targeted information. Collaborative filtering is a classic and widely used recommendation method.However, GCN-based recommendation methods still have many problems. For example, GCN recursively merges messages from different order neighborhoods, mixing different node messages indistinguishably leads to training difficulties, and problems such as over-smoothing have great constraints on the recommendation model. Different from the current GCN-based method, in order to solve the above problems, it is proposed to use a simple GCN model to separately aggregate neighbor messages in different orders for collaborative filtering, and then aggregate them in a hierarchical manner without introducing other model parameters. After that, the Dropout idea is transferred to the model, and the influence of over-smoothing is reduced by randomly discarding neighbor messages at each layer, which prevents over-fitting and improves the performance of the model.Experimental results on three data sets prove the effectiveness of the proposed model.

Key words: graph data, recommendation systems, collaborative filtering, graph convolutional networks

摘要： 图卷积网络（GCN）已应用于各领域并取得巨大成功，其中包括推荐系统。推荐系统的核心目标是推测用户偏好从而主动为用户提供有价值有针对性的消息，协同过滤是经典且广泛应用的一种推荐方法。但基于GCN的推荐方法仍存在诸多问题，如GCN递归地合并来自不同阶邻域的消息，难以区分地混合不同的节点消息导致训练困难，以及过平滑等问题对推荐模型产生了很大的约束。与目前基于GCN的方法不同，针对以上问题，提出使用简单GCN模型分别汇总不同顺序的邻域消息用于协同过滤，然后以分层方式将它们聚合，无需引入其他模型参数。之后，将Dropout思想迁移至模型中，通过在每一层随机丢弃邻居消息来减轻过平滑的影响，很好地防止了过拟合并提升了模型性能。在三个数据集上进行的实验结果证明了所提模型的有效性。

关键词: 图数据, 推荐系统, 协同过滤, 图卷积网络

ZHAI Zhengli, FENG Shu, LI Penghui. Collaborative Filtering Recommendation Model Based on Improved Graph Convolutional Network[J]. Computer Engineering and Applications, 2022, 58(17): 199-205.

翟正利, 冯舒, 李鹏辉. 改进图卷积网络的协同过滤推荐模型[J]. 计算机工程与应用, 2022, 58(17): 199-205.

References

[1] KOREN Y，BELL R，VOLINSKY C.Matrix factorization techniques for recommender systems[J].Computer，2009，42（8）：30-37.
[2] HE X N，LIAO L Z.Neural collaborative filtering[C]//Proceedings of the 26th International Conference on World Wide Web，2017：173-182.
[3] XUE H J，DAI X，ZHANG J，et al.Deep matrix factori-zation models for recommender systems[C]//Proceedings of IJCAI，2017：3203-3209.
[4] CHEN X，XU H，ZHANG Y，et al.Sequential recommendation with user memory networks[C]//Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining，2018：108-116.
[5] TANG J X，WANG K.Personalized top-n sequential recommendation via convolutional sequence embedding[C]//Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining，2018：565-573.
[6] 王永，赵旭辉，李晓光，等.一种面向协同过滤的快速最近邻居搜索方法[J].计算机工程与应用，2021，57（17）：96-105.
WANG Y，ZHAO X H，LI X G，et al.Fast nearest-neighbor searching method for collaborative filtering[J].Computer Engineering and Applications，2021，57（17）：96-105.
[7] DEFFERRARD M，BRESSON X，VANDERGHEYNST P.Convolutional neural networks on graphs with fast localized spectral filtering[C]//Advances in Neural Information Processing Systems，2016：3844-3852.
[8] SHUMAN D I，NARANG S K，FROSSARD P，et al.The emerging field of signal processing on graphs：extending high-dimensional data analysis to networks and other irregular domains[J].IEEE Signal Processing Magazine，2013，30（3）：83-98.
[9] BRUNA J，ZAREMBA W，SZLAM A，et al.Spectral networks and locally connected networks on graphs[J].arXiv：1312.6203v3，2013.
[10] CHENG Z，DING Y，HE X，et al.A3NCF：an adaptive aspect attention model for rating prediction[C]//Proceedings of IJCAI，2018：3748-3754.
[11] SAHOO N，SINGH P，MUKHOPADHYAY T.A hidden Markov model for collaborative filtering[J].MIS Quarterly，2012，36（4）：1329-1356.
[12] KOREN Y.Factorization meets the neighborhood：a multi-faceted collaborative fifiltering model[C]//Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining，2008：426-434.
[13] 邢长征，赵宏宝，张全贵，等.融合评论文本层级注意力和外积的推荐方法[J].计算机科学与探索，2020，14（6）：947-957.
XING C Z，ZHAO H B，ZHANG Q G，et al.Review text hierarchical attention and outer product for recommendation method[J].Journal of Frontiers of Computer Science and Technology，2020，14（6）：947-957.
[14] YING R，HE R，CHEN K F，et al.Graph convolutional neural networks for web-scale recommender systems[J].arXiv：1806.01973，2018.
[15] WANG X，HE X，WANG M，et al.Neural graph collaborative filtering[J].arXiv：1905.08108，2019.
[16] CHEN L，WU L，HONG R C，et al.Revisiting graph based collaborative filtering：a linear residual graph convolutional[J].arXiv：2001.10167，2020.
[17] HAMILTON W，YING Z，LESKOVEC J.Inductive representation learning on large graphs[C]//NeurIPS，2017：1025-1035.
[18] WU F，SOUZA A，ZHANG T，et al.Simplifying graph convolutional networks[C]//Proceedings of the 36th International Conference on Machine Learning，2019：6861-6871.
[19] RENDLE S，FREUDENTHALER C，GANTNER Z，et al.BPR：Bayesian personalized ranking from implicit feedback[J].arXiv：1205.2618，2012.
[20] ZHANG S，YAO L N，SUN A X，et al.NeuRec：on nonlinear transformation for personalized ranking[J].arXiv：1805.03002，2018.
[21] BERG R，KIPF T N，WELLING M.Graph convolutional matrix completion[J].arXiv：1706.02263，2017.
[22] WANG X，HE X N，WANG M，et al.Neural graph collaborative filtering[J].arXiv：1905.08108，2019.
[23] CHEN L，WU L，HONG R C，et al.Revisiting graph based collaborative filtering：a linear residual graph convolutional network approach[J].arXiv：2001.10167，2020.
[24] NT H，MAEHARA T.Revisiting graph neural networks：all we have is low-pass filters[J].arXiv：1905.09550，2019.