Evaluating Graph Generative Models with Contrastively Learned Features
This work addresses the need for better evaluation methods in graph generation, which is important for researchers and practitioners in machine learning and network science, though it is incremental as it builds on existing GNN and contrastive learning techniques.
The paper tackled the problem of evaluating graph generative models by proposing the use of contrastively trained Graph Neural Networks (GNNs) for more reliable metrics, showing that this approach outperforms traditional subgraph counting and random GNNs, with Graph Substructure Networks (GSNs) further improving the ability to distinguish distances between graph datasets.
A wide range of models have been proposed for Graph Generative Models, necessitating effective methods to evaluate their quality. So far, most techniques use either traditional metrics based on subgraph counting, or the representations of randomly initialized Graph Neural Networks (GNNs). We propose using representations from contrastively trained GNNs, rather than random GNNs, and show this gives more reliable evaluation metrics. Neither traditional approaches nor GNN-based approaches dominate the other, however: we give examples of graphs that each approach is unable to distinguish. We demonstrate that Graph Substructure Networks (GSNs), which in a way combine both approaches, are better at distinguishing the distances between graph datasets.