While retrieval-augmented generation systems partially address the hallucination issues in large language models, it also introduces new vulnerabilities to knowledge corruption attacks. Adversaries exploit these vulnerabilities by poisoning documents provided by RAG system to manipulate LLM outputs. To counter this threat, we propose SecureCollaRAG, a Byzantine-tolerant collaborative RAG framework leveraging Multi-source Knowledge Validation Mechanism. Our approach enables agent system to securely verify document provenance through dynamic GNN-based credibility scoring, effectively preventing stealthy knowledge corruption attacks while preserving essential domain knowledge integrity. Through extensive evaluations and formal analysis, we demonstrate that SecureCollaRAG maintains robustness against attackers under non-IID data distributions.

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.

Personalized federated recommendation systems can not only extract common prior knowledge from extensive decentralized data but also provide personalized models for different users to achieve independent and customized services. Incorporating user relationship graphs to enhance personalized modeling is highly promising in federated recommendation. However, it is challenging to construct such graphs and further capture personalized user information while guaranteeing multi-level (i.e., data-level and edge-level) privacy in reality. To this end, in this paper, we propose P 2 FedRec, a relationship-aware P rivacy-preserving and P ersonalized Fed erated Rec ommendation scheme, which can achieve multi-level privacy protection with personalized modeling guarantees. Specifically, we first develop a user-server collaborative mechanism for relationship graph generation and user-specific preferences capture in a privacy-preserving manner. Then, we design an embedding-shared local graph construction module and a noisy global graph-guided aggregation module to safeguard the data-level and edge-level privacy, respectively. Moreover, we introduce a personalized model training module that enables users to learn tailored local models. Theoretical analysis demonstrates that P 2 FedRec achieves both data-level and edge-level privacy preservation on the user and server sides. Extensive experiments conducted on five real-world datasets highlight the outstanding performance of P 2 FedRec.

Digital forensics is crucial to fight crimes around the world. Decentralized Digital Forensics (DDF) promotes it to another level by channeling the power of blockchain into digital investigations. In this work, we focus on the privacy and security of DDF. Our motivations arise from (1) how to track an anonymous-and-malicious data user who leaks only a part of the previously requested data, (2) how to achieve access control while protecting data from untrusted data centers, and (3) how to enable efficient and secure search on the blockchain. To address these issues, we propose Themis: an anonymous and secure DDF scheme with traceable anonymity, private access control, and efficient search. Our framework is boosted by establishing a Trusted Execution Environment in each authority (blockchain node) for securing the uploading, requesting, and searching. To instantiate the framework, we design a secure and robust watermarking scheme in conjunction with decentralized anonymous authentication, a private and fine-grained access control scheme, and an efficient and secure search scheme based on a dynamically updated data structure. We formally define and prove the privacy and security of Themis. We build a prototype with Ethereum and Intel SGX2 to evaluate its performance, which supports processing data from a considerable number of data providers and investigators.

Vertical Federated Learning (VFL) effectively addresses the issue of data isolation, which makes data mining secure. Most VFL implementations rely on a single server or third party for training, which will be terminated if the server or third party fails. In addition, the model accuracy trained by VFL depends on the quality of the client’s local features; nevertheless, the client’s local feature quality is difficult to verify. There exists a chance that the features owned by the client are irrelevant to the model or the intermediate results submitted by the client are inaccurate, such that the model’s accuracy will be seriously affected. In order to solve the single point failure and model accuracy issues in VFL, this paper first proposes a Blockchain – based Decentralized VFL (BDVFL) training model. With the integration of blockchain and the VFL training process, the nodes within the blockchain are categorized into non-training and training nodes. Our method focuses on the scenario in which all training nodes possess labeled data and actively engage in the training procedure of VFL. To be specific, first, each client utilizes local features and initial models to carry out forward activation and generate intermediate results. Second, we randomly choose a training node and combine it with the intermediate results from all clients to formulate the loss function. Finally, each client updates the local model by using the gradient. To protect the raw features, a blinding factor is utilized for safeguarding the intermediate results submitted by the client, such that the training nodes cannot infer the local features from intermediate results. To mitigate the interference of irrelevant training outcomes from clients on the model’s accuracy, we propose a verifiable aggregation method to assess the validity of the intermediate results submitted by the clients. We have conducted both theoretical and experimental analysis, and the results demonstrate the effectiveness of the proposed method.

The minimum path pair (MPP) query problem is to find the optimal meeting point of two minimum paths for two users in a network, where each user's minimum path has its own departure and destination nodes. However, the MPP query problem on a time-varying network that generally exists in the world remains open. In this study, we investigate the time-varying minimum path pair (TMPP) query problem, where the arcs of the network are time dependent. We first model the TMPP and then propose a time impulse neural network (TINN) to solve the TMPP. In the design of the TINN, the entire network topology is considered as the architecture of the neural network, and each node is viewed as a neuron. The core of an impulse-based neuron consists of six parts: input, impulse receiver, time window selector, neuron storage, impulse sender, and output. Unlike the traditional neural network, the whole neural network does not require a training process but is implemented through an impulse mechanism. The TINN consists of two stages; the first stage is to find the two minimum paths of two users, while the second stage is to calculate the distance between these two minimum paths. The underlying idea of the TINN is to find the minimum path using an impulse mechanism. The minimum path relies on the earliest time impulse stemming from the depart node that arrives at the destination node. With this mechanism, both the minimum path of the network and the distance between two paths can be addressed. Furthermore, theoretical analysis demonstrates the correctness and provides the complexity of the TINN. Experiments of the TINN are carried out based on the well-known New York City Map. A comparative study illustrates the effectiveness of the TINN.