Explainability Techniques for Graph Convolutional Networks

By Federico Baldassarre, Hossein Azizpour. International Conference on Machine Learning, 2019

335 citations recorded by the source corpus.

Graph Networks are used to make decisions in potentially complex scenarios but it is usually not obvious how or why they made them. In this work, we study the explainability of Graph Network decisions using two main classes of techniques, gradient-based and decomposition-based, on a toy dataset and a chemistry task. Our study sets the ground for future development as well as application to real-world problems.

Subfields: Artificial Intelligence, Materials Chemistry

View Paper: https://www.semanticscholar.org/paper/8fb202cdcfec3b0e7ba0e3f88949d6d923b48b2d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Shuiwang Ji
2. Haiyang Yu
3. Hao Yuan
4. Jiliang Tang
5. Siqi Miao
6. Miaoyuan Liu
7. Shurui Gui
8. Jie Wang

In Materials Chemistry

1. Muhan Zhang
2. Hao Yuan
3. Jiayun Luo
4. Pan Li
5. Weiwei Jiang
6. Jiliang Tang
7. Shuiwang Ji
8. Shurui Gui

Citations in the corpus

• Graph Neural Network for Traffic Forecasting: A Survey (2021, methodology): “The explainability study for graph neural networks in other studies is still in the early stage [248], [249], [250] and it remains a challenge how these explainability techniques can be used in the traffic forecasting domain.”
• Graph Neural Networks in Recommender Systems: A Survey (2020, background): “Themethods can be divided into two categories: the instance-levelmethods provide example-specific explanations by identifying important input features for its prediction [3, 123, 194]; the model-level methods provide high-level interpretations and a generic understanding of how deep graph models wor”
• Interpretable deep learning: interpretation, interpretability, trustworthiness, and beyond (2021, background): “Recently, more researches focus on the interpretations of GNN models, such as GraphLIME [76], CoGE [51], Counterfactual explanations on GNNs [18] and others [20, 111, 132].”
• Explainability in Graph Neural Networks: A Taxonomic Survey (2020, methodology): “[54] applies the -stabilized rule from the original LRP, while Schwarzenberg et al.”
• On Explainability of Graph Neural Networks via Subgraph Explorations (2021, methodology): “First, several methods employ gradient values or feature values to study the importance of input graph nodes, edges, or node features (Baldassarre & Azizpour, 2019; Pope et al., 2019).”
• BrainGNN: Interpretable Brain Graph Neural Network for fMRI Analysis (2020, methodology): “works have been used for explaining GNNs using post-hoc interpretation meth526 ods [35,3,53].”
• XGNN: Towards Model-Level Explanations of Graph Neural Networks (2020, background): “The existing study [4, 40] only provides example-level explanations for graph models.”
• Interpretable and Generalizable Graph Learning via Stochastic Attention Mechanism (2022, background): “Some works check the gradients w.r.t. the input features to find important features (Pope et al., 2019; Baldassarre & Azizpour, 2019).”