Generating Human Understandable Explanations for Node Embeddings
Not on the map. Why?
Abstract (arXiv)
Node embedding algorithms produce low-dimensional latent representations of nodes in a graph. These embeddings are often used for downstream tasks, such as node classification and link prediction. In this paper, we investigate the following two questions: (Q1) Can we explain each embedding dimension with human-understandable graph features (e.g. degree, clustering coefficient and PageRank). (Q2) How can we modify existing node embedding algorithms to produce embeddings that can be easily explained by human-understandable graph features? We find that the answer to Q1 is yes and introduce a new framework called XM (short for eXplain eMbedding) to answer Q2. A key aspect of XM involves minimizing the nuclear norm of the generated explanations. We show that by minimizing the nuclear norm, we minimize the lower bound on the entropy of the generated explanations. We test XM on a variety of real-world graphs and show that XM not only preserves the performance of existing node embedding methods, but also enhances their explainability.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 32 of its 56 references that are in the map. First posted on arXiv 11 Jun 2024.
Builds on
32 of 56 references.
References · 32
- 2018Deep Graph Infomaxcited by 1,101
- 2019GNNExplainer: Generating Explanations for Graph Neural Networkscited by 735
- 2015LINE: Large-scale Information Network Embeddingcited by 582
- 2019Multi-scale Attributed Node Embeddingcited by 543
- 2021On Explainability of Graph Neural Networks via Subgraph Explorationscited by 280
- 2020A Survey on the Explainability of Supervised Machine Learningcited by 225
- 2020Graph Representation Learning via Graphical Mutual Information Maximizationcited by 217
- 2020PGM-Explainer: Probabilistic Graphical Model Explanations for Graph Neural Networkscited by 210
- 2020GraphLIME: Local Interpretable Model Explanations for Graph Neural Networkscited by 209
- 2019Position-aware Graph Neural Networkscited by 203
- 2019Explainability Techniques for Graph Convolutional Networkscited by 160
- 2020Machine Learning on Graphs: A Model and Comprehensive Taxonomycited by 156
20 more
- 2020Higher-Order Explanations of Graph Neural Networks via Relevant Walkscited by 138
- 2017Cooperative Game Theory Approaches for Network Partitioningcited by 134
- 2022Evaluating Explainability for Graph Neural Networkscited by 85
- 2019Analogies Explained: Towards Understanding Word Embeddingscited by 63
- 2021GraphSVX: Shapley Value Explanations for Graph Neural Networkscited by 63
- 2018Towards Understanding Linear Word Analogiescited by 57
- 2020RelEx: A Model-Agnostic Relational Model Explainercited by 53
- 2018On the Dimensionality of Word Embeddingcited by 46
- 2017Multi-channel Weighted Nuclear Norm Minimization for Real Color Image Denoisingcited by 41
- 2019Stacking Models for Nearly Optimal Link Prediction in Complex Networkscited by 35
- 2022A Survey of Explainable Graph Neural Networks: Taxonomy and Evaluation Metricscited by 26
- 2022FlowX: Towards Explainable Graph Neural Networks via Message Flowscited by 20
- 2021A Meta-Learning Approach for Training Explainable Graph Neural Networkscited by 15
- 2020Graph Neural Networks Including Sparse Interpretabilitycited by 13
- 2021Systematic comparison of graph embedding methods in practical taskscited by 11
- 2021Node Proximity Is All You Need: Unified Structural and Positional Node and Graph Embeddingcited by 9
- 2022Classic Graph Structural Features Outperform Factorization-Based Graph Embedding Methods on Community Labelingcited by 6
- 2018Exploring the Semantic Content of Unsupervised Graph Embeddings: An Empirical Studycited by 5
- 2019On the Interpretability and Evaluation of Graph Representation Learningcited by 4
- 2023DimCL: Dimensional Contrastive Learning For Improving Self-Supervised Learningcited by 4
Built on by
Citing papers per year: 2024: 1; 2025: 0; 2026: 1.
2 papers.