Stable Multimodal Graph Unlearning via Feature-Dimension Aware Quantile Selection

TL;DR

This paper introduces FDQ, a novel framework for stable and privacy-preserving unlearning in multimodal graph neural networks, addressing over-editing issues in high-dimensional layers.

Contribution

FDQ adaptively applies quantile thresholds to sensitive high-dimensional layers, improving utility preservation and privacy in multimodal graph unlearning.

Findings

FDQ outperforms existing methods in utility preservation.

FDQ effectively defends against membership inference attacks.

Experiments on Ele-Fashion and Goodreads-NC validate robustness.

Abstract

Graph unlearning remains a critical technique for supporting privacy-preserving and sustainable multimodal graph learning. However, we observe that existing unlearning strategies tend to apply uniform parameter selection and editing across all graph neural network (GNN) layers, which is especially harmful for multimodal graphs where high-dimensional input projections encode dominant cross-modal knowledge. As a result, over-editing these sensitive layers often leads to catastrophic utility degradation after forgetting, undermining both stable learning and effective privacy protection. To address this gap, we propose FDQ, a Feature-Dimension Aware Quantile framework for multimodal graph unlearning. FDQ adaptively identifies high-dimensional input projection layers and applies more conservative, FDQ-guided quantile thresholds when constructing suppression sets, while keeping the underlying importance estimation mechanism unchanged. FDQ is seamlessly integrated with diagonal sensitivity-based parameter importance analysis to enable efficient node and edge unlearning under general forget requests. Through extensive experiments on Ele-Fashion and Goodreads-NC, we demonstrate that FDQ consistently achieves strong utility preservation while maintaining effective forgetting against membership inference attacks. Overall, FDQ offers a principled and robust solution for privacy-aware unlearning in high-dimensional multimodal graph systems.