GDLNN: Marriage of Programming Language and Neural Networks for Accurate and Easy-to-Explain Graph Classification

GDLNN combines a domain-specific programming language, called GDL, with neural networks. The main strength of GDLNN lies in its GDL layer, which generates expressive and interpretable graph representations. Since the graph representation is interpretable, existing model explanation techniques can be directly applied to explain GDLNN's predictions. Our evaluation shows that the GDL-based representation achieves high accuracy on most graph classification benchmark datasets, outperforming dominant graph learning methods such as GNNs. Applying an existing model explanation technique also yields high-quality explanations of GDLNN's predictions. Furthermore, the cost of GDLNN is low when the explanation cost is included. In graph classification, graph representation learning holds the key to success. The goal of this task is to learn useful feature representations (embeddings) for the entire graph that effectively capture its key structure and properties. These representations are utilized in various graph machine learning tasks, including graph classification. Beyond predictive performance, learning interpretable graph representations has become increasingly important, as it directly impacts model explainability, crucial in decision-critical domains such as drug discovery (Kakkad et al., 2023).