{"type":"ScholarlyArticle","title":"What Does My GNN Really Capture? On Exploring Internal GNN Representations","authors":[{"name":"Luca Veyrin-Forrer","slug":"luca-veyrin-forrer","given":"Luca","family":"Veyrin-Forrer","affiliation":"Univ Lyon , INSA Lyon , CNRS , UCBL , LIRIS , UMR5205 , FR-69621 Villeurbanne"},{"name":"Ataollah Kamal","slug":"ataollah-kamal","given":"Ataollah","family":"Kamal","affiliation":"Univ Lyon , INSA Lyon , CNRS , UCBL , LIRIS , UMR5205 , FR-69621 Villeurbanne"},{"name":"Stefan Duffner","slug":"stefan-duffner","given":"Stefan","family":"Duffner","affiliation":"Univ Lyon , INSA Lyon , CNRS , UCBL , LIRIS , UMR5205 , FR-69621 Villeurbanne"},{"name":"Marc Plantevit","slug":"marc-plantevit","given":"Marc","family":"Plantevit","affiliation":"EPITA Research and Development Laboratory (LRDE) , FR-94276 Le Kremlin-Bicêtre"},{"name":"Céline Robardet","slug":"celine-robardet","given":"Céline","family":"Robardet","affiliation":"Univ Lyon , INSA Lyon , CNRS , UCBL , LIRIS , UMR5205 , FR-69621 Villeurbanne"}],"abstract":"Graph Neural Networks (GNNs) are very efficient at classifying graphs but their internal functioning is opaque which limits their field of application. Existing methods to explain GNN focus on disclosing the relationships between input graphs and model decision. In this article, we propose a method that goes further and isolates the internal features, hidden in the network layers, that are automatically identified by the GNN and used in the decision process. We show that this method makes possible to know the parts of the input graphs used by GNN with much less bias that SOTA methods and thus to bring confidence in the decision process.","doi":"10.24963/ijcai.2022/105","url":"https://ijcai.org/proceedings/2022/105","pdf":"https://ijcai.org/proceedings/2022/0105.pdf","markdown":"https://ijcai.org/proceedings/2022/105.md","bibtex":"https://ijcai.org/proceedings/2022/bibtex/105","volume":{"slug":"2022","title":"Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2022","url":"https://ijcai.org/proceedings/2022/","year":2022,"location":"Vienna","dates":"23-29 July 2022","editors":"Luc De Raedt","isbn":"978-1-956792-00-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2022"},"section":"AI Ethics, Trust, Fairness","track":"Main Track","number":105,"pages":{"first":747,"last":752},"keywords":["Machine Learning: Sequence and Graph Learning","Data Mining: Frequent Pattern Mining","Machine Learning: Explainable/Interpretable Machine Learning","AI Ethics, Trust, Fairness: Explainability and Interpretability"],"videos":[{"label":"Video #1","url":"https://slideslive.com/38986693","length_sec":83,"provider":"slideslive"},{"label":"Video #2","url":"https://slideslive.com/38984836","length_sec":273,"provider":"slideslive"}],"references":[{"position":1,"raw":"[Auer et al., 2002] Peter Auer, Nicolò Cesa-Bianchi, and Paul Fischer. Finite-time analysis of the multiarmed bandit problem. Mach. Learn., 47(2-3):235-256, 2002.","title":"Finite-time analysis of the multiarmed bandit problem","authors":"Auer","year":"2002","venue":"Mach. Learn","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Baldassarre and Azizpour, 2019] Federico Baldassarre and Hossein Azizpour. Explainability techniques for graph convolutional networks. CoRR, abs/1905.13686, 2019.","title":"Explainability techniques for graph convolutional networks","authors":"Federico Baldassarre, Hossein Azizpour","year":"2019","venue":null,"doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Burkart and Huber, 2021] Nadia Burkart and Marco F. Hu- ber. A survey on the explainability of supervised machine learning. J. Artif. Intell. Res., 70:245-317, 2021.","title":"A survey on the explainability of supervised machine learning","authors":"Nadia Burkart, Marco F Huber","year":"2021","venue":"J. Artif. Intell. Res","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[De Bie, 2009] Tijl De Bie. Finding interesting itemsets us- ing a probabilistic model for binary databases. Technical report, Univ. Bristol, 2009.","title":"Finding interesting itemsets using a probabilistic model for binary databases","authors":"De Bie, ; Tijl, De Bie","year":"2009","venue":null,"doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[De Bie, 2011] Tijl De Bie. An information theoretic frame- work for data mining. In Chid Apté, Joydeep Ghosh, and Padhraic Smyth, editors, ACM SIGKDD, pages 564-572. ACM, 2011.","title":"An information theoretic framework for data mining","authors":"De Bie, ; Tijl, De Bie","year":"2011","venue":"ACM SIGKDD","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Duval and Malliaros, 2021] Alexandre Duval and Fragkiskos D. Malliaros. Graphsvx: Shapley value explanations for graph neural networks. In ECML PKDD 2021, pages 302-318, 2021.","title":"Graphsvx: Shapley value explanations for graph neural networks","authors":"Alexandre Duval, Fragkiskos D Malliaros","year":"2021","venue":"ECML PKDD 2021","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Huang et al., 2020] Qiang Huang, Makoto Yamada, Yuan Tian, Dinesh Singh, Dawei Yin, and Yi Chang. Graphlime: Local interpretable model explanations for graph neural networks. CoRR, abs/2001.06216, 2020.","title":"Graphlime: Local interpretable model explanations for graph neural networks","authors":"Qiang Huang, Makoto Yamada, Yuan Tian, Dinesh Singh, Dawei Yin, Yi Chang","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Kazius et al., 2005] Jeroen Kazius, Ross McGuire, and Roberta Bursi. Derivation and validation of toxicophores for mutagenicity prediction. Jour. med. chem., 48(1):312- 320, 2005.","title":"Derivation and validation of toxicophores for mutagenicity prediction","authors":"Kazius","year":"2005","venue":"Jour. med. chem","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Kipf and Welling, 2017] Thomas N. Kipf and Max Welling. Semi-supervised classification with graph convolutional networks. In 5th International Conference on Learning Representations, ICLR 2017, Toulon, France, April 24-26, 2017, Conference Track Proceedings. OpenReview.net, 2017.","title":"Semi-supervised classification with graph convolutional networks","authors":"Thomas N Kipf, Max Welling","year":"2017","venue":"5th International Conference on Learning Representations, ICLR 2017","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Luo et al., 2020] Dongsheng Luo, Wei Cheng, Dongkuan Xu, Wenchao Yu, Bo Zong, Haifeng Chen, and Xiang Zhang. Parameterized explainer for graph neural network. In Advances in Neural Information Processing Systems 33, NeurIPS 2020, 2020.","title":"Parameterized explainer for graph neural network","authors":"Dongsheng Luo, Wei Cheng, Dongkuan Xu, Wenchao Yu, Bo Zong, Haifeng Chen, Xiang Zhang","year":"2020","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Molnar, 2020] C. Molnar. Interpretable machine learning. Lulu.com, 2020.","title":"Interpretable machine learning","authors":"C Molnar","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Morris et al., 2020] Christopher Morris, Nils M. Kriege, Franka Bause, Kristian Kersting, Petra Mutzel, and Mar- ion Neumann. TUDataset. CoRR, abs/2007.08663, 2020.","title":"TUDataset. CoRR","authors":"Christopher Morris, Nils M Kriege, Franka Bause, Kristian Kersting, Petra Mutzel, Marion Neumann","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Pope et al., 2019] Phillip E. Pope, Soheil Kolouri, Moham- mad Rostami, Charles E. Martin, and Heiko Hoffmann. Explainability methods for GCN. In IEEE CVPR, pages 10772-10781, 2019.","title":"Explainability methods for GCN","authors":"Phillip E Pope, Soheil Kolouri, Mohammad Rostami, Charles E Martin, Heiko Hoffmann","year":"2019","venue":"IEEE CVPR","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Schnake et al., 2020] Thomas Schnake, Oliver Eberle, Jonas Lederer, Shinichi Nakajima, Kristof T. Schütt, Klaus-Robert Müller, and Grégoire Montavon. XAI for graphs. CoRR, abs/2006.03589, 2020.","title":"XAI for graphs","authors":"Thomas Schnake, Oliver Eberle, Jonas Lederer, Shinichi Nakajima, T Kristof, Klaus-Robert Schütt, Grégoire Müller, Montavon","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Veyrin-Forrer et al., 2021] Luca Veyrin-Forrer, Ataollah Kamal, Stefan Duffner, Marc Plantevit, and Céline Ro- bardet. On GNN explanability with activation patterns. https://hal.archives-ouvertes.fr/hal-03367714, 2021.","title":"On GNN explanability with activation patterns","authors":"Luca Veyrin-Forrer, Ataollah Kamal, Stefan Duffner, Marc Plantevit, Céline Robardet","year":"2021","venue":null,"doi":null,"url":"https://hal.archives-ouvertes.fr/hal-03367714","internal_url":null},{"position":16,"raw":"[Vu and Thai, 2020] Minh N. Vu and My T. Thai. Pgm- explainer: Probabilistic graphical model explanations for graph neural networks. In NeurIPS 2020, 2020.","title":"Pgmexplainer: Probabilistic graphical model explanations for graph neural networks","authors":"Minh N Vu, My T Thai","year":"2020","venue":"NeurIPS 2020","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Wu et al., 2017] Zhenqin Wu, Bharath Ramsundar, Evan N. Feinberg, Joseph Gomes, Caleb Geniesse, Aneesh S. Pappu, Karl Leswing, and Vijay S. Pande. Moleculenet. CoRR, abs/1703.00564, 2017.","title":"Moleculenet. CoRR","authors":"Wu","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Yan and Han, 2002] Xifeng Yan and Jiawei Han. GSPAN: Graph-based substructure pattern mining. In (ICDM 2002), pages 721-724. IEEE Computer Society, 2002.","title":"GSPAN: Graph-based substructure pattern mining","authors":"Yan, Han ; ] Xifeng Yan, Jiawei Han","year":"2002","venue":"ICDM 2002)","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Ying et al., 2019] Zhitao Ying, Dylan Bourgeois, Jiaxuan You, Marinka Zitnik, and Jure Leskovec. Gnnexplainer: Generating explanations for graph neural networks. In NeurIPS 2019, pages 9240-9251, 2019.","title":"Gnnexplainer: Generating explanations for graph neural networks","authors":"Ying","year":"2019","venue":"NeurIPS 2019","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Yuan et al., 2020a] Hao Yuan, Jiliang Tang, Xia Hu, and Shuiwang Ji. XGNN: towards model-level explanations of graph neural networks. In ACM SIGKDD, pages 430-438, 2020.","title":"XGNN: towards model-level explanations of graph neural networks","authors":"Yuan","year":"2020","venue":"ACM SIGKDD","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Yuan et al., 2020b] Hao Yuan, Haiyang Yu, Shurui Gui, and Shuiwang Ji. Explainability in graph neural networks: A taxonomic survey. CoRR, abs/2012.15445, 2020.","title":"Explainability in graph neural networks: A taxonomic survey","authors":"Yuan","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}