Graph-Fraudster: Adversarial Attacks on Graph Neural Network-Based Vertical Federated Learning (2021-10-13T00:00:00.000000Z)

TL;DR

A novel adversarial attack method, named Graph-Fraudster, which generates adversarial perturbations based on the noise-added global node embeddings via the privacy leakage and the gradient of pairwise node that can remain a threat to GVFL even if two possible defense mechanisms are applied.

Abstract

Graph neural network (GNN) has achieved great success on graph representation learning. Challenged by large-scale private data collected from user side, GNN may not be able to reflect the excellent performance, without rich features and complete adjacent relationships. Addressing the problem, vertical federated learning (VFL) is proposed to implement local data protection through training a global model collaboratively. Consequently, for graph-structured data, it is a natural idea to construct a GNN-based VFL (GVFL) framework. However, GNN has been proven vulnerable to adversarial attacks. Whether the vulnerability will be brought into the GVFL has not been studied. This is the first study of adversarial attacks on GVFL. A novel adversarial attack method is proposed, named Graph-Fraudster. It generates adversarial perturbations based on the noise-added global node embeddings via the privacy leakage and the gradient of pairwise node. Specifically, first, Graph-Fraudster steals the global node embeddings and sets up a shadow model of the server for the attack generator. Second, noise is added into node embeddings to confuse the shadow model. Finally, the gradient of pairwise node is used to generate attacks with the guidance of noise-added node embeddings. Extensive experiments on five benchmark datasets demonstrate that Graph-Fraudster achieves the state-of-the-art attack performance compared with baselines in different GNN based GVFLs. Furthermore, Graph-Fraudster can remain a threat to GVFL even if two possible defense mechanisms are applied. In addition, some suggestions are put forward for the future work to improve the robustness of GVFL. The code and datasets can be downloaded at https://github.com/hgh0545/Graph-Fraudster.

References48 items

Generative Ensemble Learning for Mitigating Adversarial Malware Detection in IoT

Federated Learning for Open Banking

A Vertical Federated Learning Framework for Graph Convolutional Network

Sustainable Security for the Internet of Things Using Artificial Intelligence Architectures

Privacy-Preserving Federated Learning on Partitioned Attributes

Label Leakage and Protection in Two-party Split Learning

GraphFL: A Federated Learning Framework for Semi-Supervised Node Classification on Graphs

ASFGNN: Automated separated-federated graph neural network

Feature Inference Attack on Model Predictions in Vertical Federated Learning

Adversarial Attack on Large Scale Graph

Vertically Federated Graph Neural Network for Privacy-Preserving Node Classification

Privacy-Preserving Graph Neural Network for Node Classification

FedSel: Federated SGD under Local Differential Privacy with Top-k Dimension Selection

FedNER: Medical Named Entity Recognition with Federated Learning

FedNER: Privacy-preserving Medical Named Entity Recognition with Federated Learning

PrivacyFL: A Simulator for Privacy-Preserving and Secure Federated Learning

Gabor-oriented local order feature-based deep learning for face annotation

A Restricted Black-Box Adversarial Framework Towards Attacking Graph Embedding Models

Robust Graph Convolutional Networks Against Adversarial Attacks

Attacking Graph Convolutional Networks via Rewiring

Deep Leakage from Gradients

Bayesian Nonparametric Federated Learning of Neural Networks

Adversarial Examples on Graph Data: Deep Insights into Attack and Defense

Simplifying Graph Convolutional Networks

Agnostic Federated Learning

APPLIED FEDERATED LEARNING: IMPROVING GOOGLE KEYBOARD QUERY SUGGESTIONS

Comprehensive Privacy Analysis of Deep Learning: Passive and Active White-box Inference Attacks against Centralized and Federated Learning

GA-Based Q-Attack on Community Detection

Attack Graph Convolutional Networks by Adding Fake Nodes

Fast Gradient Attack on Network Embedding

Adversarial Attack on Graph Structured Data

Adversarial Attacks on Neural Networks for Graph Data

Adversarial Attacks on Classification Models for Graphs

Deep Models Under the GAN: Information Leakage from Collaborative Deep Learning

Federated Optimization: Distributed Machine Learning for On-Device Intelligence

Semi-Supervised Classification with Graph Convolutional Networks

Communication-Efficient Learning of Deep Networks from Decentralized Data

Federated Learning of Deep Networks using Model Averaging

Explaining and Harnessing Adversarial Examples

Collective Classification in Network Data

Differential Privacy: A Survey of Results

The political blogosphere and the 2004 U.S. election: divided they blog

Automating the Construction of Internet Portals with Machine Learning

FedGraphNN: A Federated Learning System and Benchmark for Graph Neural Networks

Deep Feature Representation and Similarity Matrix based Noise Label Refinement Method for Efficient Face Annotation

Collective Classi!cation in Network Data

FedGNN : Federated graph neural network for privacy - preserving recommendation Deep models under the GAN : Information leakage from collaborative deep learning , ” in Proc . ACM SIGSAC Conf . Comput

Federated machine learning : Concept and applications