Toward Scalable Graph Unlearning: A Node Influence Maximization based Approach

TL;DR

This paper introduces a scalable graph unlearning framework that leverages influence maximization to improve data removal efficiency and effectiveness in large-scale graph models, addressing key scalability and entanglement challenges.

Contribution

It proposes Node Influence Maximization (NIM) and a scalable unlearning framework (SGU) that enhance unlearning performance and scalability in web-scale graph scenarios.

Findings

NIM improves unlearning forgetting capabilities.

SGU achieves state-of-the-art performance on large datasets.

Approach is scalable and adaptable to existing GU methods.

Abstract

Machine unlearning, as a pivotal technology for enhancing model robustness and data privacy, has garnered significant attention in prevalent web mining applications, especially in thriving graph-based scenarios. However, most existing graph unlearning (GU) approaches face significant challenges due to the intricate interactions among web-scale graph elements during the model training: (1) The gradient-driven node entanglement hinders the complete knowledge removal in response to unlearning requests; (2) The billion-level graph elements in the web scenarios present inevitable scalability issues. To break the above limitations, we open up a new perspective by drawing a connection between GU and conventional social influence maximization. To this end, we propose Node Influence Maximization (NIM) through the decoupled influence propagation model and fine-grained influence function in a scalable manner, which is crafted to be a plug-and-play strategy to identify potential nodes affected by unlearning entities. This approach enables offline execution independent of GU, allowing it to be seamlessly integrated into most GU methods to improve their unlearning performance. Based on this, we introduce Scalable Graph Unlearning (SGU) as a new fine-tuned framework, which balances the forgetting and reasoning capability of the unlearned model by entity-specific optimizations. Extensive experiments on 14 datasets, including large-scale ogbn-papers100M, have demonstrated the effectiveness of our approach. Specifically, NIM enhances the forgetting capability of most GU methods, while SGU achieves comprehensive SOTA performance and maintains scalability.