Disentangled and Self-Explainable Node Representation Learning

TL;DR

This paper introduces DiSeNE, an unsupervised framework for generating interpretable node embeddings by disentangling topological features, with new metrics and extensive validation on benchmark datasets.

Contribution

We propose a novel unsupervised method for learning disentangled, self-explainable node embeddings aligned with graph topology, along with new evaluation metrics.

Findings

Effective in producing interpretable embeddings

Outperforms existing methods on benchmark datasets

Provides new metrics for interpretability assessment

Abstract

Node representations, or embeddings, are low-dimensional vectors that capture node properties, typically learned through unsupervised structural similarity objectives or supervised tasks. While recent efforts have focused on explaining graph model decisions, the interpretability of unsupervised node embeddings remains underexplored. To bridge this gap, we introduce DiSeNE (Disentangled and Self-Explainable Node Embedding), a framework that generates self-explainable embeddings in an unsupervised manner. Our method employs disentangled representation learning to produce dimension-wise interpretable embeddings, where each dimension is aligned with distinct topological structure of the graph. We formalize novel desiderata for disentangled and interpretable embeddings, which drive our new objective functions, optimizing simultaneously for both interpretability and disentanglement. Additionally, we propose several new metrics to evaluate representation quality and human interpretability. Extensive experiments across multiple benchmark datasets demonstrate the effectiveness of our approach.