Joint Subgraph-to-Subgraph Transitions -- Generalizing Triadic Closure for Powerful and Interpretable Graph Modeling

TL;DR

This paper introduces the Subgraph-to-Subgraph Transition (SST) framework, generalizing triadic closure to model graph evolution, enabling interpretable link prediction with performance comparable to state-of-the-art graph neural networks.

Contribution

It proposes a novel SST framework that generalizes triadic closure, providing interpretable graph evolution models and link prediction methods validated against existing neural network approaches.

Findings

SST models achieve comparable performance to state-of-the-art GNNs.

The framework offers highly interpretable results.

Algorithms and code are provided for practical implementation.

Abstract

We generalize triadic closure, along with previous generalizations of triadic closure, under an intuitive umbrella generalization: the Subgraph-to-Subgraph Transition (SST). We present algorithms and code to model graph evolution in terms of collections of these SSTs. We then use the SST framework to create link prediction models for both static and temporal, directed and undirected graphs which produce highly interpretable results. Quantitatively, our models match out-of-the-box performance of state of the art graph neural network models, thereby validating the correctness and meaningfulness of our interpretable results.