Disentangled and Self-Explainable Node Representation Learning

2024-10-28

arXiv.org

AI Summary

Unsupervised node embedding methods lack intrinsic interpretability, hindering human understanding of what each embedding dimension represents. Method: This paper proposes DiSeNE, the first framework for dimension-level interpretable node embedding, where each dimension explicitly corresponds to a specific topological pattern (e.g., centrality, clustering, bridging). It introduces a formal joint optimization objective that simultaneously enforces disentanglement and interpretability, integrating graph topology-aware modeling with self-supervised learning. Additionally, it establishes the first evaluation metric suite designed for human-understandable interpretability of embeddings. Results: Extensive experiments on multiple benchmark datasets demonstrate that DiSeNE significantly improves both embedding disentanglement and human interpretability, consistently outperforming state-of-the-art unsupervised node embedding methods across all evaluated metrics.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics

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.

Problem

• Generating unsupervised self-explainable node embeddings
• Aligning embedding dimensions with distinct topological structures
• Simultaneously optimizing for interpretability and disentanglement objectives

Innovation

• Unsupervised disentangled representation learning for node embeddings
• Dimension-wise interpretability aligned with graph topology
• Novel objective functions optimizing interpretability and disentanglement