标签独立信息压缩的异质图表示

doi:10.3969/j.issn.1006-2475.2025.04.006

摘要/Abstract

摘要： 现有异质图（Heterogeneous Graph， HG）表示方法大都基于强大的图神经网络，聚合元路径内及元路径之间的语义信息来嵌入节点。然而，现有方法忽视了HG中节点的异质性，导致邻居节点中的无关信息沿着复杂结构扩散到高阶节点，扰动HG表示。为克服该问题，本文提出一种标签独立信息压缩的异质图表示方法LICHGR （Label-independent Information Compression for Heterogeneous Graph Representation)。LICHGR的核心思想是在信息瓶颈的指导下，利用希尔伯特-斯密特独立性准则限制异质图中标签独立信息的传播而尽可能保留标签依赖的信息。具体地，LICHGR通过在输入特征、元路径内隐藏特征、真实标签之间构造多方面的标签独立信息压缩限制，抽取丰富的标签依赖的信息，从而提高异质图表示质量。在3个公开的数据集上设计的多个实验充分验证了LICHGR的有效性。

关键词: 图神经网络, 异质图表示, 信息瓶颈, 希尔伯特-斯密特独立性准则

Abstract: The existing methods for heterogeneous graph （HG） representation are mostly based on powerful graph neural networks， which aggregate semantic information within and between meta-paths to embed nodes. However， these existing approaches overlook the heterogeneity of nodes in HG， causing irrelevant information from neighboring nodes to spread along graph structures to higher-order nodes， disturbing the HG representation. To overcome this problem， this paper proposes a heterogeneous graph representation method called Label-Independent Compression for Heterogeneous Graph （LICHGR）. The core idea of LICHGR is， under the guidance of the Information Bottleneck， to utilize the Hilbert-Schmidt Independence Criterion to restrict the propagation of label-independent information in heterogeneous graph while preserving label-dependent information as much as possible. Specifically， LICHGR constructs multi-faceted label-independent compression constraints among input features， hidden features within meta-paths， and true labels， extracting rich label-dependent information to enhance the quality of heterogeneous graph representation. Multiple experiments designed on three public datasets validate the effectiveness of LICHGR.

Key words: graph neural network, heterogeneous graph representation, information bottleneck, Hilbert-Schmidt independence criterion

中图分类号:

TP391

马剑, 王怡菲, 孟丽, 何云飞, 杨飞. 标签独立信息压缩的异质图表示[J]. 计算机与现代化, 2025, 0(04): 36-41.

MA Jian, WANG Yifei, MENG Li, HE Yunfei, YANG Fei. Label-independent Information Compression for Heterogeneous Graph Representation[J]. Computer and Modernization, 2025, 0(04): 36-41.

参考文献

［1］ SCARSELLI F， GORI M， TSOI A C， et al. The graph neural network model［J］. IEEE Transactions on Nneural Networks， 2008，20（1）： 61-80.
［2］ KIPF T N， WELLING M. Semi-supervised classification with graph convolutional network［C］// 5th International Conference on Learning Representations. ICLR， 2017. DOI:
10.48550/arXiv.1609.02907.
［3］姚春华，张学磊，宋馨宇，等.一种基于图卷积神经网络和依存分析的财经新闻情感分析方法［J］. 计算机与现代化， 2022（5）：33-39.
［4］胡鹏，阚红星，王永康，等. 基于异构图注意力网络的药物相互作用预测［J］. 中国医疗设备， 2023，38（5）：110-114.
［5］夏义春，李汪根，李豆豆，等. 结合注意力机制和图神经网络的CTR预估模型［J］. 计算机与现代化， 2023（3）：29-37.
［6］ WEI X M， LIU Y Z， SUN J S， et al. Dual subgraph-based graph neural network for friendship prediction in location-based social networks［J］. ACM Transactions on Knowledge Discovery from Data， 2023，17（3）. DOI: 10.1145/355
4981.
［7］ LI M， CAI X R， XU S H， et al. Metapath-aggregated heterogeneous graph neural network for drug-target interaction prediction［J］. Briefings in Bioinformatics， 2023，24（1）. DOI: 10.1093/bib/bbac578.
［8］ LI Z F， LIU H， ZHANG Z L， et al. Learning knowledge graph embedding with heterogeneous relation attention networks［J］. IEEE Transactions on Neural Networks and Learning Systems， 2022，33（8）：3961-3973.
［9］ WANG L， LI P P， XIONG K， et al. Modeling heterogeneous graph network on fraud detection： A community-based framework with attention mechanism［C］// Proceedings of the 30th ACM International Conference on Information & Knowledge Management. ACM， 2021：1959-1968.
［10］ HAMILTION W L， YING R， LESKOVEC J. Inductive representation learning on large graphs［C］// Proceedings of the 31st International Conference on Neural Information Processing Systems. ACM， 2017：1025-1035.
［11］ VELIC[C]KOVIC P， CUCURULL G， CASANOVA A， et al. Graph attention networks［C］// The 6th International Conference on Learning Representations. ICLR， 2018. DOI: 10.
48550/arXiv.1710.10903.
［12］ DONG X R， ZHANG Y J， PANG K， et al. Heterogeneous graph neural networks with denoising for graph embeddings［J］. Knowledge-Based Systems， 2022，238（C）. DOI: 10.1016
/j.knosys.2021.107899.
［13］ WU T L， REN H Y， LI P， et al. Graph information bottleneck［C］// Procedings of te 34th International Conference on Neural Information Processing Systems. ACM， 2020：20437-20448.
［14］ XU K Y L， HU W H， LESKOVEC J， et al. How powerful are graph neural networks?［C］// The 7th International Conference on Learning Representations. ICLR， 2019. DOI:10.
48550/arXiv.1810.00826.
［15］ WANG X， JI H Y， SHI C， et al. Heterogeneous graph attention network［C］// Proceedings of the World Wide Web Conference 2019. ACM， 2019：2022-2032.
［16］ FANG Y， LIN W Q， ZHENG V W， et al. Metagraph-based learning on heterogeneous graphs［J］. IEEE Transactions on Knowledge and Data Engineering， 2019，33（1）：154-168.
［17］ WANG Z H， YU D H， LI Q， et al. SR-HGN： Semantic-and relation-aware heterogeneous graph neural network［J］. Expert Systems with Applications， 2023， 224（C）. DOI: 10.
1016/j.eswa.2023.119982.
［18］ JI H Y， WANG X， SHI C， et al. Heterogeneous graph propagation network［J］. IEEE Transactions on Knowledge and Data Engineering， 2021，35（1）：521-532.
［19］ FU X Y， ZHANG J N， MENG Z Q， et al. MAGNN： Metapath aggregated graph neural network for heterogeneous graph embedding［C］// Proceedings of the Web Conference 2020. ACM， 2020：2331-2341.
［20］ TISHBY N， ZASLAVSKY N. Deep learning and the information bottleneck principle［C］// Proceedings of the 2015 IEEE Information Theory Workshop. IEEE， 2015. DOI: 10.
1109/ITW.2015.7133169.
［21］ GRETTON A， BOUSQUET O， SMOLA A， et al. Measuring statistical dependence with Hilbert-Schmidt norms［C］// Proceedings of the 16th International Conference on Algorithmic Learning Theory. Springer， 2005：63-77.
［22］ MA W D K， LEWIS J P， KLEIJN W B. The HSIC bottleneck： Deep learning without back-propagation［C］// Proceedings of the 34th AAAI Conference on Artificial Intelligence. AAAI， 2020，34（4）：5085-5092.
［23］ HE Y F， ZHANG Y W， QI L Y， et al. Outer product enhanced heterogeneous information network embedding for Recommendation［J］. Expert Systems with Applications， 2021，169. DOI: 10.1016/j.eswa.2020.114359.
［24］ GREENFELD D， SHALIT U. Robust learning with the Hilbert-Schmidt independence criterion［C］// Proceedings of the 37th International Conference on Machine Learning. ACM， 2020，352：3759-3768.
［25］ HE Y F， YAN D C， XIE W X， et al. Optimizing graph neural network with multiaspect Hilbert-Schmidt independence criterion［J］. IEEE Transactions on Neural Networks and Learning Systems， 2023，34（12）：10775-10788.
［26］ HUANG Q， YAMADA M， TIAN Y， et al. GraphLIME： Local interpretable model explanations for graph neural networks［J］. IEEE Transactions on Knowledge and Data Engineering， 2023，35（7）：6968-6972.
［27］ WANG M J， ZHENG D， YE Z H， et al. Deep graph library：A graph-centric，highly-performant package for graph neural networks［J］. arXiv preprint arXiv：1909.01315， 2019.