Graph neural networks (GNNs) with differential privacy (DP) offer a reliable solution for safeguarding sensitive information within graph data. Nonetheless, existing DP-based privacy-preserving GNN learning frameworks generally overlook the local topological heterogeneity of graph nodes and tailor the same privacy budget for all nodes, which may lead to either overprotection or underprotection of some nodes, potentially diminishing model utility or posing privacy leakage risks. To address this issue, we propose a Topology-aware Differential Privacy Graph Neural Network learning framework, termed TDP-GNN, which can achieve personalized privacy protection for each node with improved privacy-utility guarantees. Specifically, TDP-GNN first identifies the topological importance of each node via an adjacency information entropy method. Then, the personalized topology-aware privacy budget is designed to quantify the privacy sensitivity of each node and adaptively allocate the privacy protection strength. Besides, a weighted neighborhood aggregation mechanism is proposed during the message-passing process of GNN training, which can eliminate the impact of the introduced differentiated DP noise on the utility of the GNN model. Since TDP-GNN is based on node-level local DP, it can be seamlessly integrated into any GNN architecture in a plug-and-play manner while ensuring formal privacy guarantees. Theoretical analysis indicates that TDP-GNN achieves ε-differential privacy over the entire graph nodes while providing personalized privacy protection. Extensive experiments demonstrate that TDP-GNN consistently yields better utilities when applied to various GNN architectures (e.g., GCN and GraphSAGE) across a diverse set of benchmarks.

Graph neural networks (GNNs) have been widely applied in various graph analysis tasks. To provide more convenient and faster predictive services, many enterprises are choosing to deploy GNNs in cloud environments. However, given the increasing privacy concerns about GNNs models and graph data, as well as the need to quickly generate embeddings for new nodes in real-world applications, a critical issue in this emerging paradigm is to ensure the security and scalability of GNN predictions. In this paper, we propose a privacy-preserving and scalable GNN prediction scheme, named PS-GNN, to address the privacy issues in cloud environments. Specifically, PS-GNN utilizes a customized array structure to store graph data and employs secret sharing to preserve the confidentiality of both the GNN model and graph data. Besides, the scalability of PS-GNN is achieved by aggregating feature information from local node neighborhoods in parallel. Through a detailed analysis, we demonstrate the security of PS-GNN. Extensive experiments on real-world datasets demonstrate that PS-GNN outperforms existing schemes in terms of computational and communication overhead, and reaches state-of-the-art performance on large graphs.

In cloud-based health monitoring services, healthcare centers often outsource support vector machine (SVM)-based clinical decision models to provide remote users with clinical decisions. During service provisioning, authorized external organizations like insurance companies aim to verify decision correctness to prevent fraudulent medical reimbursements. However, existing verifiable and secure SVM classification schemes have predominantly focused on user self-verification, thereby introducing potential risks of privacy leakage (such as input data exposure) in publicly verifiable scenarios. To address the aforementioned limitation, we propose a publicly verifiable and secure SVM classification scheme (PVSSVM) for cloud-based health monitoring services in a malicious setting, which can accommodate the verification needs of users or authorized external organizations with respect to potential malicious results returned by cloud servers. Specifically, we utilize homomorphic encryption and secret sharing to protect the model and data confidentiality in the cloud server, respectively. Based on a multiserver verifiable computation framework, PVSSVM achieves public verification of predicted results. Additionally, we further investigate its performance. Experimental evaluations demonstrate that PVSSVM outperforms existing state-of-the-art solutions in terms of computation and communication overhead. Notably, in the verification scenario of large-scale predictions, the proposed scheme achieves a reduction of approximately 83.71% in computation overhead through batch verification, as compared to one-by-one verification.

With the proliferation of machine learning, the cloud server has been employed to collect massive data and train machine learning models. Several privacy-preserving machine learning schemes have been suggested recently to guarantee data and model privacy in the cloud. However, these schemes either mandate the involvement of the data owner in model training or utilize high-cost cryptographic techniques, resulting in excessive computational and communication overheads. Furthermore, none of the existing work considers the malicious behavior of the cloud server during model training. In this paper, we propose the first privacy-preserving and verifiable support vector machine training scheme by employing a two-cloud platform. Specifically, based on the homomorphic verification tag, we design a verification mechanism to enable verifiable machine learning training. Meanwhile, to improve the efficiency of model training, we combine homomorphic encryption and data perturbation to design an efficient multiplication operation for the encryption domain. A rigorous theoretical analysis demonstrates the security and reliability of our scheme. The experimental results indicate that our scheme can reduce computational and communication overheads by at least 43.94% and 99.58%, respectively, compared to state-of-the-art SVM training methods.