A Survey of Explainable Graph Neural Networks: Taxonomy and Evaluation Metrics
It addresses the interpretability problem for researchers and practitioners using GNNs, but is incremental as it synthesizes existing work rather than introducing novel methods.
This survey tackles the lack of a holistic review for explainable graph neural networks (GNNs) by providing a comprehensive taxonomy and evaluation metrics for existing explanation techniques, without presenting new experimental results.
Graph neural networks (GNNs) have demonstrated a significant boost in prediction performance on graph data. At the same time, the predictions made by these models are often hard to interpret. In that regard, many efforts have been made to explain the prediction mechanisms of these models from perspectives such as GNNExplainer, XGNN and PGExplainer. Although such works present systematic frameworks to interpret GNNs, a holistic review for explainable GNNs is unavailable. In this survey, we present a comprehensive review of explainability techniques developed for GNNs. We focus on explainable graph neural networks and categorize them based on the use of explainable methods. We further provide the common performance metrics for GNNs explanations and point out several future research directions.