Counterfactual Explanation (CE) techniques have garnered attention as a means to provide insights to the users engaging with AI systems. While extensively researched in domains such as medical imaging and autonomous vehicles, Graph Counterfactual Explanation (GCE) methods have been comparatively under-explored.
GCEs generate a new graph akin to the original one, having a different outcome grounded on the underlying predictive model. Among these GCE techniques, those rooted in generative mechanisms have received relatively limited investigation, despite demonstrating impressive accomplishments in other domains, such as artistic styles and natural language modelling.
The preference for generative explainers stems from their capacity to generate counterfactual instances during inference, leveraging autonomously acquired perturbations of the input graph. Motivated by the rationales above, our study introduces RSGG-CE, a novel Robust Stochastic Graph Generator for Counterfactual Explanations able to produce counterfactual examples from the learned latent space considering a partially ordered generation sequence.
Furthermore, we undertake both quantitative and qualitative analyses to compare RSGG-CE’s performance against SoA generative explainers, highlighting its increased abilities in engendering plausible counterfactual candidates.
Figure 1: Qualitative illustration of the counterfactuals produced by RSGG-CE. Each element in the illustration represents the adjacency matrix of an instances. We colour with black those edges that are maintained as in the original instance; with green those that are added; and with red those that are removed. Blank spots are those instances for which the explainer does not produce a valid counterfactual.
Preprinted paper (including supplementary materials) available at
Final Version of the paper available at https://ojs.aaai.org/index.php/AAAI/article/view/30149
Please cite us using:
@inproceedings{Prado-Romero_Prenkaj_Stilo_2024,
title={Robust Stochastic Graph Generator for Counterfactual Explanations},
author={Prado-Romero, Mario Alfonso and Prenkaj, Bardh and Stilo, Giovanni},
year={2024},
month={Mar.},
journal={Proceedings of the AAAI Conference on Artificial Intelligence},
publisher = {AAAI Press},
DOI={10.1609/aaai.v38i19.30149},
booktitle = {Proceedings of the Thirty-Eighth AAAI Conference on Artificial Intelligence and Thirty-Sixth Conference on Innovative Applications of Artificial Intelligence and Thirteenth Symposium on Educational Advances in Artificial Intelligence},
numpages = {9},
volume={38},
number={19},
pages={21518-21526},
series = {AAAI'24/IAAI'24/EAAI'24},
url={https://ojs.aaai.org/index.php/AAAI/article/view/30149}
}
Poster presented at AAAI 2024
Slides presented at AAAI 2024
Slides can be downloaded HERE