[1]常新功,王金珏.基于图卷积集成的网络表示学习[J].智能系统学报,2022,17(3):547-555.[doi:10.11992/tis.202107048]
CHANG Xingong,WANG Jinjue.Network representation learning using graph convolution ensemble[J].CAAI Transactions on Intelligent Systems,2022,17(3):547-555.[doi:10.11992/tis.202107048]
点击复制
CHANG Xingong,WANG Jinjue.Network representation learning using graph convolution ensemble[J].CAAI Transactions on Intelligent Systems,2022,17(3):547-555.[doi:10.11992/tis.202107048]
基于图卷积集成的网络表示学习
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
17
期数:
2022年第3期
页码:
547-555
栏目:
学术论文—智能系统
出版日期:
2022-05-05
- Title:
- Network representation learning using graph convolution ensemble
- Keywords:
- Network represents learning; Ensemble learning; Graph convolution network; Social network; Deep learning; Feature learning; Node embedding; Information network embedding
- 分类号:
- TP301.6
- 摘要:
- 针对现有网络表示学习方法泛化能力较弱等问题，提出了将stacking集成思想应用于网络表示学习的方法，旨在提升网络表示性能。首先，将3个经典的浅层网络表示学习方法DeepWalk、Node2Vec、Line作为并列的初级学习器，训练得到三部分的节点嵌入拼接后作为新数据集；然后，选择图卷积网络(graph convolutional network, GCN)作为次级学习器对新数据集和网络结构进行stacking集成得到最终的节点嵌入，GCN处理半监督分类问题有很好的效果，因为网络表示学习具有无监督性，所以利用网络的一阶邻近性设计损失函数；最后，设计评价指标分别评价初级学习器和集成后的节点嵌入。实验表明，选用GCN集成的效果良好，各评价指标平均提升了1.47~2.97倍。
- Abstract:
- Aimed at the weak generalization ability of the existing network representation learning methods, this paper proposes a stacking ensemble method that is applied to network representation learning to improve the performance of network representation. Three classical shallow network representation learning methods, namely, DeepWalk, Node2Vec, and Line, are used initially as parallel primary learners. After training, the embedded representation and three spliced node parts are obtained. Following this, the graph convolutional network (GCN) is selected as the secondary learner to stack and integrate the network structure and the new data set. This is done to obtain the final node embedded representation. GCN efficiently deals with semisupervised classification problems. As network representation learning is unsupervised, the first-order proximity of the network has been used to design the loss function. Finally, evaluation indices are designed to test the primary learner and integrated node eigenvector representation. The experimental results show that after the introduction of the integration method, each index improves by approximately 1.47–2.97 times.
- 参考文献/References:
-
[1] BHAGAT S, CORMODE G, MUTHUKRISHNAN S. Node Classification in Social Networks[M]// Social Network Data Analytics. Boston, MA: Springer, 2011: 115–148.
[2] LIBEN-NOWELL D, KLEINBERG J. The link-prediction problem for social networks[J]. Journal of the American society for information science and technology, 2007, 58(7): 1019–1031.
[3] CO?KUN M, KOYUTüRK M. Node similarity based graph convolution for link prediction in biological networks[J]. Bioinformatics (Oxford, England), 2021, 37(23): 4501–4508.
[4] DER MAATEN L V, HINTON G. Visualizing data using t-SNE[J]. Journal of machine learning research, 2008: 2579–2605.
[5] TANG Jian, LIU Jingzhou, ZHANG Ming, et al. Visualizing Large-Scale and High-Dimensional Data[C]//Proceedings of the 25th International Conference ompanion on World Wide Web. Canada, New York, 2016: 287–297.
[6] ZHANG Daokun, YIN Jie, ZHU Xingquan, et al. Network representation learning: a survey[J]. IEEE transactions on big data, 2020, 6(1): 3–28.
[7] BENGIO Y, COURVILLE A, VINCENT P. Representation learning: a review and new perspectives[J]. IEEE transactions on pattern analysis and machine intelligence, 2013, 35(8): 1798–1828.
[8] 尹赢, 吉立新, 黄瑞阳, 等. 网络表示学习的研究与发展[J]. 网络与信息安全学报, 2019, 5(2): 77–87
YIN Ying, JI Lixin, HUANG Ruiyang, et al. Research and development of network representation learning[J]. Chinese journal of network and information security, 2019, 5(2): 77–87
[9] ROWEIS S T, SAUL L K. Nonlinear dimensionality reduction by locally linear embedding[J]. Science, 2000, 290(5500): 2323–2326.
[10] HE Xiaofei, Niyogi P. Locality preserving projections[J]. In advances in neural information processing systems, 2004, 16: 153–160.
[11] TU Cunchao, ZHANG Weicheng, LIU Zhiyuan, et al. Max-margin DeepWalk: discriminative learning of network representation[C]//Proceedings of the 25th Inter-national Joint Conference on Artifificial Intelligence. New York: ACM, 2016: 3889–3895.
[12] CAO Shaosheng, LU Wei, XU Qiongkai. Grarep: Lear- ning graph representations with global structural informat- ion[C]// Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. Melbourne, Australia, 2015: 891–900.
[13] PEROZZI B, AL-RFOU R, SKIENA S. DeepWalk: online learning of social representations[C]//Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York : ACM, 2014: 701–710.
[14] GROVER A, LESKOVEC J. node2vec: scalable feature learning for networks[EB/OL]. (2016–07–03)[ 2021–07–23]https://arxiv.org/abs/1607.00653.
[15] TANG Jian, QU Meng, WANG Mingzhe, et al. LINE: large-scale information network embedding[C]//Procee- dings of the 24th International Conference on World Wide Web. New York: ACM, 2015: 1067–1077.
[16] WANG Daixin, CUI Peng, ZHU Wenwu. Structural deep network embedding[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, USA, 2016: 1225–1234.
[17] HAMILTON W, YING Z, LESKOVEC J. Inductive repres- entation learning on large graphs[C]//Advances in Neural Information Processing Systems. Long Beach , USA, 2017: 1024–1034.
[18] WANG Hongwei, WANG Jia , WANG Jialin, et al. Graph- GAN: graph representation learning with generative adve- rsarial nets[C]// Proceedings of the 32th AAAI Conference on Artificial Intelligence, New Orleans, USA, 2018: 2508–2515.
[19] ZHANG Boyu, IANG Ji, WANG Xin. Network representation learning with ensemble methods[J]. Neur- ocomputing, 2020, 380: 141–149.
[20] 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 Bingbing, CEN Keting, HUANG Junjie, et al. A survey on graph convolutional neural network[J]. Chinese journal of computers, 2020, 43(5): 755–780
[21] KIPF T N, WELLING M . Semi-supervised classifificati- on with graph convolutional networks[EB/OL]. (2016–09–09)[ 2021–07–23]https://arxiv.org/abs/1609.02907.
[22] 周志华. 机器学习[M]. 北京: 清华大学出版社, 2016.
[23] MIKOLOV T, SUTSKEVER I, CHEN KAI, et al. Distributed representations of words and phrases and their compositionality[EB/OL]. (2013–19–16)[ 2021–07–23]https://arxiv.org/abs/1310.4546v1.
[24] SUN ZHIQING, DENG ZHI-HONG, NIE JIAN-YUN, et al. RotatE: knowledge graph embedding by relational rotation in complex space[EB/OL]. (2019–02–26)[ 2021–07-23]https://arxiv.org/abs/1902.10197v1.
- 相似文献/References:
-
[1]胡小生,温菊屏,钟勇.动态平衡采样的不平衡数据集成分类方法[J].智能系统学报,2016,11(2):257.[doi:10.11992/tis.201507015]
HU Xiaosheng,WEN Juping,ZHONG Yong.Imbalanced data ensemble classification using dynamic balance sampling[J].CAAI Transactions on Intelligent Systems,2016,11():257.[doi:10.11992/tis.201507015]
[2]闵帆,王宏杰,刘福伦,等.SUCE:基于聚类集成的半监督二分类方法[J].智能系统学报,2018,13(6):974.[doi:10.11992/tis.201711027]
MIN Fan,WANG Hongjie,LIU Fulun,et al.SUCE: semi-supervised binary classification based on clustering ensemble[J].CAAI Transactions on Intelligent Systems,2018,13():974.[doi:10.11992/tis.201711027]
[3]张本才,王志海,孙艳歌.一种多样性和精度加权的数据流集成分类算法[J].智能系统学报,2019,14(1):179.[doi:10.11992/tis.201806021]
ZHANG Bencai,WANG Zhihai,SUN Yange.An ensemble classification algorithm based on diversity and accuracy weighting for data streams[J].CAAI Transactions on Intelligent Systems,2019,14():179.[doi:10.11992/tis.201806021]
[4]常亮,张伟涛,古天龙,等.知识图谱的推荐系统综述[J].智能系统学报,2019,14(2):207.[doi:10.11992/tis.201805001]
CHANG Liang,ZHANG Weitao,GU Tianlong,et al.Review of recommendation systems based on knowledge graph[J].CAAI Transactions on Intelligent Systems,2019,14():207.[doi:10.11992/tis.201805001]
[5]张燕,杜红乐.基于异构距离的集成分类算法研究[J].智能系统学报,2019,14(4):733.[doi:10.11992/tis.201807023]
ZHANG Yan,DU Hongle.Imbalanced heterogeneous data ensemble classification based on HVDM-KNN[J].CAAI Transactions on Intelligent Systems,2019,14():733.[doi:10.11992/tis.201807023]
[6]张潇鲲,刘琰,陈静.引入外部词向量的文本信息网络表示学习[J].智能系统学报,2019,14(5):1056.[doi:10.11992/tis.201809037]
ZHANG Xiaokun,LIU Yan,CHEN Jing.Representation learning using network embedding based on external word vectors[J].CAAI Transactions on Intelligent Systems,2019,14():1056.[doi:10.11992/tis.201809037]
[7]张蕾,钱峰,赵姝,等.基于多粒度结构的网络表示学习[J].智能系统学报,2019,14(6):1233.[doi:10.11992/tis.201905045]
ZHANG Lei,QIAN Feng,ZHAO Shu,et al.Network representation learning based on multi-granularity structure[J].CAAI Transactions on Intelligent Systems,2019,14():1233.[doi:10.11992/tis.201905045]
[8]贾中浩,宾辰忠,古天龙,等.基于知识图谱和用户长短期偏好的个性化景点推荐[J].智能系统学报,2020,15(5):990.[doi:10.11992/tis.201904064]
JIA Zhonghao,BIN Chenzhong,GU Tianlong,et al.Personalized attraction recommendation based on the knowledge graph and users’ long-term and short-term preferences[J].CAAI Transactions on Intelligent Systems,2020,15():990.[doi:10.11992/tis.201904064]
[9]闫涵,张旭秀,张净丹.多感知兴趣区域特征融合的图像识别方法[J].智能系统学报,2021,16(2):263.[doi:10.11992/tis.201906032]
YAN Han,ZHANG Xuxiu,ZHANG Jingdan.Image recognition method based on multi-perceptual interest region feature fusion[J].CAAI Transactions on Intelligent Systems,2021,16():263.[doi:10.11992/tis.201906032]
[10]汤礼颖,贺利乐,何林,等.一种卷积神经网络集成的多样性度量方法[J].智能系统学报,2021,16(6):1030.[doi:10.11992/tis.202011023]
TANG Liying,HE Lile,HE Lin,et al.Diversity measuring method of a convolutional neural network ensemble[J].CAAI Transactions on Intelligent Systems,2021,16():1030.[doi:10.11992/tis.202011023]
备注/Memo
收稿日期:2021-07-23。
基金项目:国家自然科学基金项目（61906110）；山西财经大学研究生创新项目（21cxxj088）.
作者简介:常新功，教授，博士，CCF高级会员，主要研究方向为图神经网络、数据挖掘、进化算法。主持10项山西省重点课题。发表学术论文30余篇;王金珏，硕士研究生，主要研究方向为图神经网络、数据挖掘
通讯作者:常新功.E-mail：c_x_g@126.com
更新日期/Last Update:
1900-01-01