Graph Unlearning Meets Influence-aware Negative Preference Optimization

2025-10-22

AI Summary

To address the sharp utility degradation of graph neural networks (GNNs) upon node removal, this paper proposes Influence-aware Negative Preference Optimization (I-NPO). I-NPO first estimates edge influence to identify high-impact edges, then designs an influence-aware message-passing mechanism and introduces a topological entropy loss to mitigate local structural information loss. By integrating negative preference optimization with entropy regularization, it effectively suppresses gradient explosion and enhances forgetting robustness. Experiments on five real-world graph datasets demonstrate that I-NPO consistently outperforms state-of-the-art methods in forgetting quality—measured by metrics such as forgetting rate and retained-set accuracy—while significantly preserving model utility. The implementation is publicly available.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Natural Language Processing: Learning & Optimization for NLP
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Recent advancements in graph unlearning models have enhanced model utility by preserving the node representation essentially invariant, while using gradient ascent on the forget set to achieve unlearning. However, this approach causes a drastic degradation in model utility during the unlearning process due to the rapid divergence speed of gradient ascent. In this paper, we introduce extbf{INPO}, an extbf{I}nfluence-aware extbf{N}egative extbf{P}reference extbf{O}ptimization framework that focuses on slowing the divergence speed and improving the robustness of the model utility to the unlearning process. Specifically, we first analyze that NPO has slower divergence speed and theoretically propose that unlearning high-influence edges can reduce impact of unlearning. We design an influence-aware message function to amplify the influence of unlearned edges and mitigate the tight topological coupling between the forget set and the retain set. The influence of each edge is quickly estimated by a removal-based method. Additionally, we propose a topological entropy loss from the perspective of topology to avoid excessive information loss in the local structure during unlearning. Extensive experiments conducted on five real-world datasets demonstrate that INPO-based model achieves state-of-the-art performance on all forget quality metrics while maintaining the model's utility. Codes are available at href{https://github.com/sh-qiangchen/INPO}{https://github.com/sh-qiangchen/INPO}.

Problem

• Slowing gradient ascent divergence in graph unlearning
• Reducing unlearning impact by targeting high-influence edges
• Maintaining model utility while achieving effective forget quality

Innovation

• Influence-aware Negative Preference Optimization for graph unlearning
• Amplifies influence of unlearned edges with message function
• Uses topological entropy loss to preserve local structure