Probing GNN Explainers: A Rigorous Theoretical and Empirical Analysis of GNN Explanation Methods

Chirag Agarwal, Marinka Zitnik, Himabindu Lakkaraju
Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, PMLR 151:8969-8996, 2022.

Abstract

As Graph Neural Networks (GNNs) are increasingly being employed in critical real-world applications, several methods have been proposed in recent literature to explain the predictions of these models. However, there has been little to no work on systematically analyzing the reliability of these methods. Here, we introduce the first-ever theoretical analysis of the reliability of state-of-the-art GNN explanation methods. More specifically, we theoretically analyze the behavior of various state-of-the-art GNN explanation methods with respect to several desirable properties (e.g., faithfulness, stability, and fairness preservation) and establish upper bounds on the violation of these properties. We also empirically validate our theoretical results using extensive experimentation with nine real-world graph datasets. Our empirical results further shed light on several interesting insights about the behavior of state-of-the-art GNN explanation methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v151-agarwal22b, title = { Probing GNN Explainers: A Rigorous Theoretical and Empirical Analysis of GNN Explanation Methods }, author = {Agarwal, Chirag and Zitnik, Marinka and Lakkaraju, Himabindu}, booktitle = {Proceedings of The 25th International Conference on Artificial Intelligence and Statistics}, pages = {8969--8996}, year = {2022}, editor = {Camps-Valls, Gustau and Ruiz, Francisco J. R. and Valera, Isabel}, volume = {151}, series = {Proceedings of Machine Learning Research}, month = {28--30 Mar}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v151/agarwal22b/agarwal22b.pdf}, url = {https://proceedings.mlr.press/v151/agarwal22b.html}, abstract = { As Graph Neural Networks (GNNs) are increasingly being employed in critical real-world applications, several methods have been proposed in recent literature to explain the predictions of these models. However, there has been little to no work on systematically analyzing the reliability of these methods. Here, we introduce the first-ever theoretical analysis of the reliability of state-of-the-art GNN explanation methods. More specifically, we theoretically analyze the behavior of various state-of-the-art GNN explanation methods with respect to several desirable properties (e.g., faithfulness, stability, and fairness preservation) and establish upper bounds on the violation of these properties. We also empirically validate our theoretical results using extensive experimentation with nine real-world graph datasets. Our empirical results further shed light on several interesting insights about the behavior of state-of-the-art GNN explanation methods. } }
Endnote
%0 Conference Paper %T Probing GNN Explainers: A Rigorous Theoretical and Empirical Analysis of GNN Explanation Methods %A Chirag Agarwal %A Marinka Zitnik %A Himabindu Lakkaraju %B Proceedings of The 25th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2022 %E Gustau Camps-Valls %E Francisco J. R. Ruiz %E Isabel Valera %F pmlr-v151-agarwal22b %I PMLR %P 8969--8996 %U https://proceedings.mlr.press/v151/agarwal22b.html %V 151 %X As Graph Neural Networks (GNNs) are increasingly being employed in critical real-world applications, several methods have been proposed in recent literature to explain the predictions of these models. However, there has been little to no work on systematically analyzing the reliability of these methods. Here, we introduce the first-ever theoretical analysis of the reliability of state-of-the-art GNN explanation methods. More specifically, we theoretically analyze the behavior of various state-of-the-art GNN explanation methods with respect to several desirable properties (e.g., faithfulness, stability, and fairness preservation) and establish upper bounds on the violation of these properties. We also empirically validate our theoretical results using extensive experimentation with nine real-world graph datasets. Our empirical results further shed light on several interesting insights about the behavior of state-of-the-art GNN explanation methods.
APA
Agarwal, C., Zitnik, M. & Lakkaraju, H.. (2022). Probing GNN Explainers: A Rigorous Theoretical and Empirical Analysis of GNN Explanation Methods . Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 151:8969-8996 Available from https://proceedings.mlr.press/v151/agarwal22b.html.

Related Material