Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.04807 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917190133874688 |
|---|---|
| author | Llorente, Oscar Boal, Jaime Sánchez-Úbeda, Eugenio F. Diaz-Cano, Antonio Familiar, Miguel |
| author_facet | Llorente, Oscar Boal, Jaime Sánchez-Úbeda, Eugenio F. Diaz-Cano, Antonio Familiar, Miguel |
| contents | Graph Neural Networks (GNNs) have demonstrated remarkable performance in a wide range of tasks, such as node classification, link prediction, and graph classification, by exploiting the structural information in graph-structured data. However, in node classification, computing node-level explainability becomes extremely time-consuming as the size of the graph increases, while batching strategies often degrade explanation quality. This paper introduces a novel approach to parallelizing node-level explainability in GNNs through graph partitioning. By decomposing the graph into disjoint subgraphs, we enable parallel computation of explainability for node neighbors, significantly improving the scalability and efficiency without affecting the correctness of the results, provided sufficient memory is available. For scenarios where memory is limited, we further propose a dropout-based reconstruction mechanism that offers a controllable trade-off between memory usage and explanation fidelity. Experimental results on real-world datasets demonstrate substantial speedups, enabling scalable and transparent explainability for large-scale GNN models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_04807 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Parallelizing Node-Level Explainability in Graph Neural Networks Llorente, Oscar Boal, Jaime Sánchez-Úbeda, Eugenio F. Diaz-Cano, Antonio Familiar, Miguel Machine Learning Artificial Intelligence Graph Neural Networks (GNNs) have demonstrated remarkable performance in a wide range of tasks, such as node classification, link prediction, and graph classification, by exploiting the structural information in graph-structured data. However, in node classification, computing node-level explainability becomes extremely time-consuming as the size of the graph increases, while batching strategies often degrade explanation quality. This paper introduces a novel approach to parallelizing node-level explainability in GNNs through graph partitioning. By decomposing the graph into disjoint subgraphs, we enable parallel computation of explainability for node neighbors, significantly improving the scalability and efficiency without affecting the correctness of the results, provided sufficient memory is available. For scenarios where memory is limited, we further propose a dropout-based reconstruction mechanism that offers a controllable trade-off between memory usage and explanation fidelity. Experimental results on real-world datasets demonstrate substantial speedups, enabling scalable and transparent explainability for large-scale GNN models. |
| title | Parallelizing Node-Level Explainability in Graph Neural Networks |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2601.04807 |