Is Meta-Path Attention an Explanation? Evidence of Alignment and Decoupling in Heterogeneous GNNs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Jiang, Maiqi, Ali, Noman, Ding, Yiran, Zhang, Yanfu
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914315610619904
author Jiang, Maiqi
Ali, Noman
Ding, Yiran
Zhang, Yanfu
author_facet Jiang, Maiqi
Ali, Noman
Ding, Yiran
Zhang, Yanfu
contents Meta-path-based heterogeneous graph neural networks aggregate over meta-path-induced views, and their semantic-level attention over meta-path channels is widely used as a narrative for ``which semantics matter.'' We study this assumption empirically by asking: when does meta-path attention reflect meta-path importance, and when can it decouple? A key challenge is that most post-hoc GNN explainers are designed for homogeneous graphs, and naive adaptations to heterogeneous neighborhoods can mix semantics and confound perturbations. To enable a controlled empirical analysis, we introduce MetaXplain, a meta-path-aware post-hoc explanation protocol that applies existing explainers in the native meta-path view domain via (i) view-factorized explanations, (ii) schema-valid channel-wise perturbations, and (iii) fusion-aware attribution, without modifying the underlying predictor. We benchmark representative gradient-, perturbation-, and Shapley-style explainers on ACM, DBLP, and IMDB with HAN and HAN-GCN, comparing against xPath and type-matched random baselines under standard faithfulness metrics. To quantify attention reliability, we propose Meta-Path Attention--Explanation Alignment (MP-AEA), which measures rank correlation between learned attention weights and explanation-derived meta-path contribution scores across random runs. Our results show that meta-path-aware explanations typically outperform random controls, while MP-AEA reveals both high-alignment and statistically significant decoupling regimes depending on the dataset and backbone; moreover, retraining on explanation-induced subgraphs often preserves, and in some noisy regimes improves, predictive performance, suggesting an explanation-as-denoising effect.
format Preprint
id arxiv_https___arxiv_org_abs_2602_08500
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Is Meta-Path Attention an Explanation? Evidence of Alignment and Decoupling in Heterogeneous GNNs
Jiang, Maiqi
Ali, Noman
Ding, Yiran
Zhang, Yanfu
Machine Learning
Meta-path-based heterogeneous graph neural networks aggregate over meta-path-induced views, and their semantic-level attention over meta-path channels is widely used as a narrative for ``which semantics matter.'' We study this assumption empirically by asking: when does meta-path attention reflect meta-path importance, and when can it decouple? A key challenge is that most post-hoc GNN explainers are designed for homogeneous graphs, and naive adaptations to heterogeneous neighborhoods can mix semantics and confound perturbations. To enable a controlled empirical analysis, we introduce MetaXplain, a meta-path-aware post-hoc explanation protocol that applies existing explainers in the native meta-path view domain via (i) view-factorized explanations, (ii) schema-valid channel-wise perturbations, and (iii) fusion-aware attribution, without modifying the underlying predictor. We benchmark representative gradient-, perturbation-, and Shapley-style explainers on ACM, DBLP, and IMDB with HAN and HAN-GCN, comparing against xPath and type-matched random baselines under standard faithfulness metrics. To quantify attention reliability, we propose Meta-Path Attention--Explanation Alignment (MP-AEA), which measures rank correlation between learned attention weights and explanation-derived meta-path contribution scores across random runs. Our results show that meta-path-aware explanations typically outperform random controls, while MP-AEA reveals both high-alignment and statistically significant decoupling regimes depending on the dataset and backbone; moreover, retraining on explanation-induced subgraphs often preserves, and in some noisy regimes improves, predictive performance, suggesting an explanation-as-denoising effect.
title Is Meta-Path Attention an Explanation? Evidence of Alignment and Decoupling in Heterogeneous GNNs
topic Machine Learning
url https://arxiv.org/abs/2602.08500