Structure and Feature Fusion Graph Hierarchical Pooling Model

doi:10.3778/j.issn.1673-9418.2104082

Abstract

Abstract: Graph neural networks (GNN), which extend deep neural networks to non-Euclidean data, have been proven to be powerful for numerous graph related tasks such as node classification, link prediction, and graph classification. On the task of graph classification, recent studies aim to learn graph-level representation through a hierarchical poo-ling procedure using the local and global structure information of the graph. After comparing and analyzing the cur-rent graph classification model, this paper proposes the structure and feature fusion pooling model (SAFPool) consi-dering shortcomings of the current method and combining the advantages of different methods. SAFPool utilizes two assignment matrix generation modules during the pooling process, which are structure-based cluster learning and feature-based cluster learning modules. Structure-based cluster learning module clusters nodes with similar struc-tures based on graph structure information, and feature-based cluster learning clusters nodes with similar features based on graph node features. Then the two clustering assignment matrices are weighted and aggregated to obtain the clustering assignment matric which implements the clustering strategy to utilize graph structure and node feature information at the same time. Finally, comparative experiments and visualization on multiple graph classification datasets demonstrate the effectiveness of using graph node information and structure information to implement a clustering strategy in graph classification.

Key words: graph neural network, graph classification, graph pooling, clustering assignment matrix, hierarchical model

摘要： 作为深度神经网络向非欧式数据上的扩展，图神经网络（GNN）已经在图节点分类任务、链接预测任务和图分类任务中取得了显著成就。在图分类任务上，当前方法一般通过层次化的池化过程同时考虑图的局部和全局结构信息以学习高层次的图表示。在对当前的图分类模型进行对比分析后，考虑当前方法的不足，结合不同方法的优势，提出结构和特征融合池化模型（SAFPool）。SAFPool模型在池化时使用了两个聚类分配矩阵生成模块，分别是基于结构的聚类学习和基于特征的聚类学习模块，基于结构的聚类学习根据图结构信息对结构相似的节点聚类，基于特征的聚类学习则根据图节点特征对特征相似的节点聚类。二者的聚类结果加权聚合后便能获取实现聚类策略的聚类分配矩阵以同时利用图结构和节点特征信息。最后，在多个图分类数据集上通过对比实验和可视化说明了同时显式地利用图节点特征信息和图结构信息实现聚类策略的有效性。

关键词: 图神经网络, 图分类, 图池化, 聚类分配矩阵, 层次化模型

MA Fuyuan, WANG Ying, LI Lina, WANG Hongji. Structure and Feature Fusion Graph Hierarchical Pooling Model[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(1): 179-186.

马涪元, 王英, 李丽娜, 汪洪吉. 融合结构和特征的图层次化池化模型[J]. 计算机科学与探索, 2023, 17(1): 179-186.

References

[1] ZHANG L, WANG X D, LI H S, et al. Structure-feature based graph self-adaptive pooling[C]//Proceedings of the Web Con-ference 2020, Taipei, China, Apr 20-24, 2020. New York: ACM, 2020: 3098-3104.
[2] ZHANG Z, BU J J, MARTIN E, et al. Hierarchical graph pooling with structure learning[J]. arXiv:1911.05954, 2019.
[3] BRUNA J, ZAREMBA W, SZLAM A, et al. Spectral net-works and locally connected networks on graphs[C]//Procee-dings of the 2nd International Conference on Learning Representations, Banff, Apr 14-16, 2014: 1-14.
[4] DEFFERRARD M, BRESSON X, VANDERGHEYNST P. Convolutional neural networks on graphs with fast loca-lized spectral fifiltering[C]//Proceedings of the 30th Inter-national Conference on Neural Information Processing Sys-tems, Dec 5-10, 2016. Red Hook: Curran Associates, 2016:3844-3852.
[5] KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks[C]//Proceedings of the 5th International Conference on Learning Representations, Toulon, Apr 24-26, 2017: 1-14.
[6] GILMER J, SCHOENHOLZ S S, Riley P F, et al. Neural message passing for quantum chemistry[C]//Proceedings of the 34th International Conference on Machine Learning, Sydney, Aug 6-11, 2017: 1263-1272.
[7] HAMILTON W L, YING Z T, LESKOVEC J. Inductive representation learning on large graphs[C]//Proceedings of the Annual Conference on Neural Information Processing Systems 2017, Long Beach, Dec 4-9, 2017. Red Hook: Curran Associates, 2017: 1024-1034.
[8] VELI?KOVI? P, CUCURULL G, CASANOVA A, et al. Graph attention networks[C]//Proceedings of the 2018 Interna-tional Conference on Learning Representations, Vancouver, Apr 30-May 3, 2018: 1-12.
[9] XU K, HU W H, LESKOVEC J, et al. How powerful are graph neural networks?[C]//Proceedings of the 7th Interna-tional Conference on Learning Representations, New Orleans, May 6-9, 2019: 1-17.
[10] YING Z T, YOU J X, MORRIS C, et al. Hierarchical graph representation learning with differentiable pooling[C]//Procee-dings of the Annual Conference on Neural Information Pro-cessing Systems 2018, Montréal, Dec 3-8, 2018. Red Hook: Curran Associates, 2018: 4805-4815.
[11] MA Y, WANG S H, AGGARWAL C C, et al. Graph convo-lutional networks with EigenPooling[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowle-dge Discovery and Data Mining, Anchorage, Aug 4-8, 2019. New York: ACM, 2019: 723-731.
[12] YUAN H, JI S W. StructPool: structured graph pooling via conditional random fields[C]//Proceedings of the 8th Inter-national Conference on Learning Representations, Addis Ababa, Apr 26-30, 2020: 1-12.
[13] GAO H Y, JI S W. Graph U-Nets[C]//Proceedings of the 36th International Conference on Machine Learning, Long Beach, Jun 9-15, 2019: 2083-2092.
[14] LEE J, LEE I, KANG J. Self-attention graph pooling[C]//Proceedings of the 36th International Conference on Machine Learning, Long Beach, Jun 9-15, 2019: 3734-3743.
[15] CHENG J P, DONG L, LAPATA M. Long short-term memory-networks for machine reading[C]//Proceedings of the 2016 Conference on Empirical Methods in Natural Language Pro-cessing, Austin, Nov 1-4, 2016. Stroudsburg: ACL, 2016: 551-561.
[16] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Proceedings of the Annual Conference on Neural Information Processing Systems 2017, Long Beach, Dec 4-9, 2017. Red Hook: Curran Associates, 2017: 5998-6008.
[17] BORGWARDT K M, ONG C S, SCH?NAUER S, et al. Protein function prediction via graph kernels[C]//Procee-dings of the 13th International Conference on Intelligent Systems for Molecular Biology 2005, Detroit, Jun 25-29, 2005: 47-56.
[18] SHERVASHIDZE N, SCHWEITZER P, LEEUWEN E J, et al. Weisfeiler-Lehman graph kernels[J]. Journal of Machine Learning Research, 2011, 12: 2539-2561.
[19] LI W, WANG G G, GANDOMI A H. A survey of learning-based intelligent optimization algorithms[J]. Archives of Com-putational Methods in Engineering, 2021, 28(5): 3781-3799.
[20] WANG G G, GANDOMI A H, ALAVI A H, et al. A com-prehensive review of krill herd algorithm: variants, hybrids and applications[J]. Artificial Intelligence Review, 2019, 51(1): 119-148.
[21] WANG G G, SUASH D, CUI Z H. Monarch butterfly opti-mization[J]. Neural Computing and Applications, 2019, 31(7): 1995-2014.
[22] WANG G G, DEB S, COELHO L D S. Earthworm opti-misation algorithm: a bio-inspired metaheuristic algorithm for global optimisation problems[J]. International Journal of Bio-Inspired Computation, 2018, 12(1): 1-22.
[23] WANG G G. Moth search algorithm: a bio-inspired meta-heuristic algorithm for global optimization problems[J]. Meme-tic Computing, 2018, 10(2): 151-164.