The Interpretable and Effective Graph Neural Additive Networks

TL;DR

This paper introduces GNAN, an interpretable graph neural network model based on additive models, which provides transparent explanations while maintaining competitive accuracy with traditional black-box GNNs.

Contribution

The paper presents GNAN, a novel interpretable GNN model that offers visual explanations and maintains high accuracy, addressing transparency issues in graph learning.

Findings

GNAN achieves accuracy comparable to black-box GNNs.

GNAN provides clear visual explanations at feature and graph levels.

The model is effective across various tasks and datasets.

Abstract

Graph Neural Networks (GNNs) have emerged as the predominant approach for learning over graph-structured data. However, most GNNs operate as black-box models and require post-hoc explanations, which may not suffice in high-stakes scenarios where transparency is crucial. In this paper, we present a GNN that is interpretable by design. Our model, Graph Neural Additive Network (GNAN), is a novel extension of the interpretable class of Generalized Additive Models, and can be visualized and fully understood by humans. GNAN is designed to be fully interpretable, offering both global and local explanations at the feature and graph levels through direct visualization of the model. These visualizations describe exactly how the model uses the relationships between the target variable, the features, and the graph. We demonstrate the intelligibility of GNANs in a series of examples on different tasks and datasets. In addition, we show that the accuracy of GNAN is on par with black-box GNNs, making it suitable for critical applications where transparency is essential, alongside high accuracy.