DT+GNN: A Fully Explainable Graph Neural Network using Decision Trees
This provides an interpretable alternative to opaque GNNs for researchers and practitioners in graph-based machine learning, though it is incremental in combining existing concepts.
The authors tackled the problem of black-box graph neural networks (GNNs) by proposing DT+GNN, a fully explainable architecture that uses decision trees for interpretability, and demonstrated it performs as well as traditional GNNs on real-world and synthetic datasets.
We propose the fully explainable Decision Tree Graph Neural Network (DT+GNN) architecture. In contrast to existing black-box GNNs and post-hoc explanation methods, the reasoning of DT+GNN can be inspected at every step. To achieve this, we first construct a differentiable GNN layer, which uses a categorical state space for nodes and messages. This allows us to convert the trained MLPs in the GNN into decision trees. These trees are pruned using our newly proposed method to ensure they are small and easy to interpret. We can also use the decision trees to compute traditional explanations. We demonstrate on both real-world datasets and synthetic GNN explainability benchmarks that this architecture works as well as traditional GNNs. Furthermore, we leverage the explainability of DT+GNNs to find interesting insights into many of these datasets, with some surprising results. We also provide an interactive web tool to inspect DT+GNN's decision making.