Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
144 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

View-based Explanations for Graph Neural Networks (2401.02086v2)

Published 4 Jan 2024 in cs.LG and cs.DB

Abstract: Generating explanations for graph neural networks (GNNs) has been studied to understand their behavior in analytical tasks such as graph classification. Existing approaches aim to understand the overall results of GNNs rather than providing explanations for specific class labels of interest, and may return explanation structures that are hard to access, nor directly queryable.We propose GVEX, a novel paradigm that generates Graph Views for EXplanation. (1) We design a two-tier explanation structure called explanation views. An explanation view consists of a set of graph patterns and a set of induced explanation subgraphs. Given a database G of multiple graphs and a specific class label l assigned by a GNN-based classifier M, it concisely describes the fraction of G that best explains why l is assigned by M. (2) We propose quality measures and formulate an optimization problem to compute optimal explanation views for GNN explanation. We show that the problem is $\Sigma2_P$-hard. (3) We present two algorithms. The first one follows an explain-and-summarize strategy that first generates high-quality explanation subgraphs which best explain GNNs in terms of feature influence maximization, and then performs a summarization step to generate patterns. We show that this strategy provides an approximation ratio of 1/2. Our second algorithm performs a single-pass to an input node stream in batches to incrementally maintain explanation views, having an anytime quality guarantee of 1/4 approximation. Using real-world benchmark data, we experimentally demonstrate the effectiveness, efficiency, and scalability of GVEX. Through case studies, we showcase the practical applications of GVEX.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (76)
  1. 2024. Code and datasets. Tingyang Chen, Dazhuo Qiu, Yinghui Wu, Arijit Khan, Xiangyu Ke, and Yunjun Gao. https://github.com/ZJU-DAILY/GVEX.
  2. Zahi Ajami and Sara Cohen. 2019. Enumerating minimal weight set covers. In IEEE International Conference on Data Engineering (ICDE). 518–529.
  3. Monte Carlo methods in PageRank computation: When one iteration is sufficient. SIAM J. Numer. Anal. 45, 2 (2007), 890–904.
  4. Robust counterfactual explanations on graph neural networks. Advances in Neural Information Processing Systems (NeurIPS) 34 (2021), 5644–5655.
  5. Protein function prediction via graph kernels. Bioinformatics 21, suppl_1 (2005), i47–i56.
  6. The effects of example-based explanations in a machine learning interface. In Proceedings of the 24th International Conference on Intelligent User Interfaces. 258–262.
  7. Maximizing a monotone submodular function subject to a matroid constraint. SIAM J. Comput. 40, 6 (2011), 1740–1766.
  8. On the red-blue set cover problem. In ACM-SIAM Symposium on Discrete Algorithms (SODA). 345–353.
  9. Amit Chakrabarti and Sagar Kale. 2015. Submodular maximization meets streaming: matchings, matroids, and more. Mathematical Programming 154 (2015), 225–247.
  10. GQP: A framework for scalable and effective graph query-based pricing. In IEEE International Conference on Data Engineering (ICDE). 1573–1585.
  11. Friendship and mobility: user movement in location-based social networks. In ACM International Conference on Knowledge Discovery and Data Mining (KDD). 1082–1090.
  12. Edith Cohen and David D Lewis. 1999. Approximating matrix multiplication for pattern recognition tasks. Journal of Algorithms 30, 2 (1999), 211–252.
  13. Enyan Dai and Suhang Wang. 2022. Towards prototype-based self-explainable graph neural network. arXiv preprint arXiv:2210.01974 (2022).
  14. Fairness in streaming submodular maximization: Algorithms and hardness. Advances in Neural Information Processing Systems (NeurIPS) 33 (2020), 13609–13622.
  15. Incremental graph pattern matching. ACM Transactions on Database Systems (TODS) 38, 3 (2013), 1–47.
  16. Answering graph pattern queries using views. In IEEE International Conference on Data Engineering (ICDE). 184–195.
  17. Induced subgraph isomorphism: Are some patterns substantially easier than others? Theoretical Computer Science 605 (2015), 119–128.
  18. A large-scale database for graph representation learning. In Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks 1, NeurIPS Datasets and Benchmarks.
  19. Efficient graph neural network inference at large scale. arXiv preprint arXiv:2211.00495 (2022).
  20. Allennlp: A deep semantic natural language processing platform. arXiv preprint arXiv:1803.07640 (2018).
  21. Stephan Günnemann. 2022. Graph neural networks: Adversarial robustness. Graph Neural Networks: Foundations, Frontiers, and Applications (2022), 149–176.
  22. Inductive representation learning on large graphs. In Advances in Neural Information Processing Systems (NeurIPS). 1024–1034.
  23. Turboiso: Towards ultrafast and robust subgraph isomorphism search in large graph databases. In ACM International Conference on Management of Data (SIGMOD). 337–348.
  24. OGB-LSC: A large-scale challenge for machine learning on graphs. In Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks 1, NeurIPS Datasets and Benchmarks.
  25. Open graph benchmark: Datasets for machine learning on graphs. Advances in neural information processing systems 33 (2020), 22118–22133.
  26. Open graph benchmark: Datasets for machine learning on graphs. In Advances in Neural Information Processing Systems (NeurIPS).
  27. Graphlime: Local interpretable model explanations for graph neural networks. IEEE Transactions on Knowledge and Data Engineering (2022).
  28. Global counterfactual explainer for graph neural networks. In ACM International Conference on Web Search and Data Mining (WSDM). 141–149.
  29. ASAP: Fast, approximate graph pattern mining at scale. In 13th USENIX Symposium on Operating Systems Design and Implementation (OSDI 18). 745–761.
  30. Drug–target affinity prediction using graph neural network and contact maps. RSC advances 10, 35 (2020), 20701–20712.
  31. Accelerating training and inference of graph neural networks with fast sampling and pipelining. Machine Learning and Systems (MLSys) 4 (2022), 172–189.
  32. Derivation and validation of toxicophores for mutagenicity prediction. Journal of Medicinal Chemistry 48, 1 (2005), 312–320.
  33. Turboflux: A fast continuous subgraph matching system for streaming graph data. In ACM International Conference on Management of Data (SIGMOD). 411–426.
  34. Diederik P Kingma and Jimmy Ba. 2015. Adam: A method for stochastic optimization. In International Conference on Learning Representations (ICLR).
  35. Thomas N Kipf and Max Welling. 2017. Semi-supervised classification with graph convolutional networks. In International Conference on Learning Representations (ICLR).
  36. Predict then propagate: Graph neural networks meet personalized pageRank. In International Conference on Learning Representations (ICLR).
  37. Discovering patterns for fact checking in knowledge graphs. Journal of Data and Information Quality (JDIQ) 11, 3 (2019), 1–27.
  38. Identifying insufficient data coverage in databases with multiple relations. Proc. VLDB Endow. 13, 11 (2020).
  39. Parameterized explainer for graph neural network. Advances in Neural Information Processing Systems (NeurIPS) 33 (2020), 19620–19631.
  40. Subgraph query generation with fairness and diversity constraints. In IEEE International Conference on Data Engineering (ICDE). 3106–3118.
  41. Imene Mami and Zohra Bellahsene. 2012. A survey of view selection methods. ACM SIGMOD Record 41, 1 (2012), 20–29.
  42. Sayan Ranu and Ambuj K Singh. 2009. Graphsig: A scalable approach to mining significant subgraphs in large graph databases. In 2009 IEEE 25th International Conference on Data Engineering. IEEE, 844–855.
  43. Marcus Schaefer and Christopher Umans. 2002. Completeness in the polynomial-time hierarchy: A compendium. SIGACT news 33, 3 (2002), 32–49.
  44. Interpreting graph neural networks for nlp with differentiable edge masking. In International Conference on Learning Representations (ICLR).
  45. Layerwise relevance visualization in convolutional text graph classifiers. In Workshop on Graph-Based Methods for Natural Language Processing (TextGraphs@EMNLP). 58–62.
  46. Taming verification hardness: an efficient algorithm for testing subgraph isomorphism. Proc. VLDB Endow. 1, 1 (2008), 364–375.
  47. Mining summaries for knowledge graph search. IEEE Transactions on Knowledge and Data Engineering 30, 10 (2018), 1887–1900.
  48. Discriminative frequent subgraph mining with optimality guarantees. Statistical Analysis and Data Mining: The ASA Data Science Journal 3, 5 (2010), 302–318.
  49. Graph attention networks. arXiv preprint arXiv:1710.10903 (2017).
  50. Counterfactual explanations for machine learning: Challenges revisited. arXiv preprint arXiv:2106.07756 (2021).
  51. Neural architecture search for GNN-based graph classification. ACM Transactions on Information Systems (2023).
  52. Graph neural networks in recommender systems: a survey. Comput. Surveys 55, 5 (2022), 1–37.
  53. A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems 32, 1 (2020), 4–24.
  54. Graph neural networks for automated de novo drug design. Drug Discovery Today 26, 6 (2021), 1382–1393.
  55. How powerful are graph neural networks?. In International Conference on Learning Representations (ICLR).
  56. Representation learning on graphs with jumping knowledge networks. In International Conference on Machine Learning (ICML). 5449–5458.
  57. Global concept-based interpretability for graph neural networks via neuron analysis. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 37. 10675–10683.
  58. Mining significant graph patterns by leap search. In Proceedings of the 2008 ACM SIGMOD International Conference on Management of Data. 433–444.
  59. Xifeng Yan and Jiawei Han. 2002. gspan: Graph-based substructure pattern mining. In IEEE International Conference on Data Mining (ICDM). 721–724.
  60. Pinar Yanardag and SVN Vishwanathan. 2015. Deep graph kernels. In ACM International Conference on Knowledge Discovery and Data Mining (KDD). 1365–1374.
  61. Graph convolutional networks for text classification. In AAAI Conference on Artificial Intelligence, Vol. 33. 7370–7377.
  62. Gnnexplainer: Generating explanations for graph neural networks. Advances in Neural Information Processing Systems (NeurIPS) 32 (2019).
  63. Hierarchical graph representation learning with differentiable pooling. Advances in Neural Information Processing Systems (NeurIPS) 31 (2018).
  64. Graph convolutional policy network for goal-directed molecular graph generation. Advances in Neural Information Processing Systems (NeurIPS) 31 (2018).
  65. Xgnn: Towards model-level explanations of graph neural networks. In ACM international conference on knowledge discovery and data mining (KDD). 430–438.
  66. Explainability in graph neural networks: A taxonomic survey. IEEE Transactions on Pattern Analysis and Machine Intelligence 45, 5 (2023), 5782–5799.
  67. On explainability of graph neural networks via subgraph explorations. In International Conference on Machine Learning (ICML). 12241–12252.
  68. Matthew D Zeiler and Rob Fergus. 2014. Visualizing and understanding convolutional networks. In Computer Vision (ECCV). 818–833.
  69. Automatic view selection in graph databases. In International Conference on Scientific and Statistical Database Management (SSDBM). 197–202.
  70. Muhan Zhang and Yixin Chen. 2018. Link prediction based on graph neural networks. Advances in Neural Information Processing Systems (NeurIPS) 31 (2018).
  71. An end-to-end deep learning architecture for graph classification. In AAAI Conference on Artificial Intelligence, Vol. 32.
  72. GStarX: Explaining graph neural networks with structure-aware cooperative games. In Advances in Neural Information Processing Systems (NeurIPS).
  73. Distributed time-respecting flow graph pattern matching on temporal graphs. World Wide Web 23 (2020), 609–630.
  74. Grain: Improving data efficiency of graph neural networks via diversified influence maximization. Proc. VLDB Endow. 14, 11 (2021), 2473–2482.
  75. Accelerating large scale real-time GNN inference using channel pruning. Proc. VLDB Endow. 14, 9 (2021).
  76. Modeling polypharmacy side effects with graph convolutional networks. Bioinformatics 34, 13 (2018), i457–i466.
Citations (7)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now