Federated learning (FL) allows multiple participants to collaboratively build deep learning (DL) models without directly sharing data. Consequently, the issue of copyright protection in FL becomes important since unreliable participants may gain access to the jointly trained model. Application of homomorphic encryption (HE) in a secure FL framework prevents the central server from accessing plaintext models. Thus, it is no longer feasible to embed the watermark at the central server using existing watermarking schemes. In this article, we propose a novel client-side FL watermarking scheme to tackle the copyright protection issue in secure FL with HE. To the best of our knowledge, it is the first scheme to embed the watermark to models under a secure FL environment. We design a black-box watermarking scheme based on client-side backdooring to embed a pre-designed trigger set into an FL model by a gradient-enhanced embedding method. Additionally, we propose a trigger set construction mechanism to ensure that the watermark cannot be forged. Experimental results demonstrate that our proposed scheme delivers outstanding protection performance and robustness against various watermark removal attacks and ambiguity attack.

When enjoying mobile crowdsensing (MCS), it is vital to evaluate the trustworthiness of mobile users (MUs) without disclosing their sensitive information. However, the existing schemes ignore this requirement in the multiple crowdsourcers (CSs) scenario. The lack of a credible sharing about MUs’ trustworthiness results in an inaccurate trust evaluation, disabling allocating tasks to reliable MUs. To address it, based on the analysis of the desired properties, we propose a scheme enablingdependabledata collection withmultiple crowdsourcerstrustsharing ($D^{2}MTS$). Specifically, we design the MU anonymous management. Two kinds of MU generated pseudonym systems without relationships are presented to mark each MU in trust evaluation and task execution, respectively. Through the devised pseudonym changes on these pseudonyms and the common token distribution algorithm,$D^{2}MTS$realizes privacy-preserving trust sharing. Moreover, to guarantee credible sharing, based on the hash chain,$D^{2}MTS$records MUs’ trustworthiness with the unforgeable signature on the blockchain established by multiple CSs which do not trust each other naturally. Extensive experiments show that compared with the other works,$D^{2}MTS$'s detection ratio of vicious MUs and the percentage of reliable MUs among the selected ones can increase by 208.61% and 28.27%. Both computational and communication delays are limited.

Federated Learning (FL) suffers from low convergence and significant accuracy loss due to local biases caused by non-Independent and Identically Distributed (non-IID) data. To enhance the non-IID FL performance, a straightforward idea is to leverage the Generative Adversarial Network (GAN) to mitigate local biases using synthesized samples. Unfortunately, existing GAN-based solutions have inherent limitations, which do not support non-IID data and even compromise user privacy. To tackle the above issues, we propose a GAN-based unbiased FL scheme, calledFlGan, to mitigate local biases using synthesized samples generated by GAN while preserving user-level privacy in the FL setting. Specifically,FlGanfirst presents a federated GAN algorithm using the divide-and-conquer strategy that eliminates the problem of model collapse in non-IID settings. To guarantee user-level privacy,FlGanthen exploits Fully Homomorphic Encryption (FHE) to design the privacy-preserving GAN augmentation method for the unbiased FL. Extensive experiments show thatFlGanachieves unbiased FL with$10\%-60\%$accuracy improvement compared with two state-of-the-art FL baselines (i.e., FedAvg and FedSGD) trained under different non-IID settings. The FHE-based privacy guarantees only cost about 0.53% of the total overhead inFlGan.

There is substantial attention to federated learning with its ability to train a powerful global model collaboratively while protecting data privacy. Despite its many advantages, federated learning is vulnerable to backdoor attacks, where an adversary injects malicious weights into the global model, making the global model's targeted predictions incorrect. Existing defenses based on identifying and eliminating malicious weights ignore the similarity variation of the local weights during iterations in the malicious model detection and the presence of benign weights in the malicious model during the malicious local weight elimination, resulting in a poor defense and a degradation of global model accuracy. In this paper, we defend against backdoor attacks from the perspective of local models. First, a malicious model detection method based on interpretability techniques is proposed. The method appends a sampling check after clustering to identify malicious models accurately. We further design a malicious local weight elimination method based on local weight contributions. This method preserves the benign weights in the malicious model to maintain their contributions to the global model. Finally, we analyze the security of the proposed method in terms of model closeness and then verify the effectiveness of the proposed method through experiments. In comparison with existing defenses, the results show that BADFL improves the global model accuracy by 23.14% while reducing the attack success rate to 0.04% in the best case.

To obtain reliable results from conflicting data in mobile crowdsensing, numerous truth discovery protocols have been proposed in the past decade. However, most of them do not consider the data privacy of entities involved (e.g., workers and servers), and several existing privacy-preserving truth discovery protocols either provide limited privacy protection or have heavy computation and communication overheads due to iterative computation and transmission over large ciphertexts.In this paper, we aim to propose privacy-preserving and lightweight truth discovery protocols to tackle the above problems. Specifically, we carefully design an anonymization protocol named AnonymTD to delink workers from their data, where workers’ data are computed and transmitted without complicated encryption. To further reduce each worker’s overheads in the scenarios where workers are willing to share their weights, we resort to the perturbation technology to propose a more lightweight truth discovery protocol named PerturbTD. Based on workers’ perturbed data, two cloud servers in PerturbTD complete most of the workload of truth discovery together, which avoids the frequent involvement of workers. The theoretical analysis and the comparative experiments in this paper demonstrate that our two protocols can achieve our security goals with low computation and communication overheads.

Community structure is a typical characteristic of complex networks. Finding communities in complex networks has many important applications, such as the advertisement and recommendation based on social networks and the discovery of new protein molecules in biological networks, which make it a hot topic in the field of complex network analysis. With the increasing concerns about the leakage of personal privacy, discovering communities spread across the local networks owned by multiple participants accurately while preserving each participant’s privacy has become an emerging challenge in distributed community detection. In this article, we propose a general federated graph learning model for privacy-preserving distributed graph learning and develop two federated clique percolation algorithms (CPAs) based on it to discover overlapping communities distributed across multiple participants’ local networks without disclosing any participant’s network privacy. Homomorphic encryption and hash operation are used in combination to protect the privacy of the vertices and edges of each local network. Furthermore, vertex attributes are involved in the calculation of clique similarity and clique percolation when dealing with attributed networks. The experimental results on real-world and artificial datasets demonstrate that the proposed algorithms achieve identical results to those of their stand-alone counterparts and more than 200% higher accuracy than the simple distributed CPAs without federating learning.

Social trust assessment that characterizes a pairwise trustworthiness relationship can spur diversified applications. Extensive efforts have been put in exploration, but mainly focusing on applying graph convolutional network to establish a social trust evaluation model, overlooking user feature factors related to context-aware information on social trust prediction. In this article, we aim to design a new trust assessment framework GATrust which integrates multi-aspect properties of users, including user context-specific information, network topological structure information, and locally-generated social trust relationships. GATrust can assigns different attention coefficients to multi-aspect properties of users in online social networks, for improving the prediction accuracy of social trust evaluation. The framework can then learn multiple latent factors of each trustor-trustee pair to establish a social trust evaluation model, by fusing graph attention network and graph convolution network. We conduct extensive experiments on two popular real-world datasets and the results exhibit that our proposed framework can improve the precision of social trust prediction, outperforming the state-of-the-art in the literature by 4.3% and 5.5% on both two datasets, respectively.

With increasingly popular GPS-equipped mobile devices (e.g., smartphones, tablets, laptops), massive spatio-textual data has been outsourced to cloud servers for storage and analysis such as spatial keyword search. However, existing privacy-preserving spatial keyword query schemes only support coarse-grained non-temporal access control in single-user sharing scenarios, which does not scale well in time-related scenes such as message valid period. To solve the above issues, we propose Privacy-preserving Boolean Range Query with Temporal access control in mobile computing (PBRQ-T). Specifically, we first achieve PBRQ with linear search complexity using the adapted Gray code, Bloom filter, and Katz-Sahai-Waters encryption. Then, we provide fine-grained and temporal access control in PBRQ based on the forward/backward derivation function and attribute-based encryption, where PBRQ is executed only when the spatio-textual data is accessible. Finally, an enhanced PBRQ-T (i.e., PBRQ-T+) with faster-than-linear search complexity is proposed by constructing a Quadtree index structure. Our formal security analysis shows that data privacy and index privacy can be guaranteed during the query process. Our extensive experiments using a real-world dataset demonstrate the efficiency and feasibility of our schemes.

To ensure the security of outsourced data without affecting data availability, one can use Symmetric Searchable Encryption (SSE) to achieve search over encrypted data. Considering that query users may search with misspelled words, the fuzzy search should be supported. However, conventional privacy-preserving fuzzy multi-keyword search schemes are incapable of achieving the result verification and adaptive security. To solve the above challenging issues, in this paper we propose a Verifiable Fuzzy multi-keyword Search scheme with Adaptive security (VFSA). VFSA first employs the locality sensitive hashing to hash the misspelled and correct keywords to the same positions, then designs a twin Bloom filter for each document to store and mask all keywords contained in the document, next constructs an index tree based on the graph-based keyword partition algorithm to achieve adaptive sublinear retrieval, finally combines the Merkle hash tree structure with the adapted multiset accumulator to check the correctness and completeness of search results. Our formal security analysis shows that VFSA is secure under the IND-CKA2 model and achieves query authentication. Our empirical experiments using the real-world dataset demonstrate the practicality of VFSA.

In recent years, the appearance of graph convolutional networks (GCNs) provides a new idea for graph structure data processing. Because of that, they can learn excellent user and item embedding by using cooperative signals of high-order neighbors, and the GCNs technique shows great potential in the recommendation. The common problem with the bulk of GCN-based models is that it appears the situation of performance degradation during the stacking of network layers. The recently proposed IMP-GCN alleviates this problem to some extent. It aims to avoid the influence of downside information from high-order propagation on embedding learning. However, we consider that it ignores the multi-interest factor, in which users may have different interests. In this paper, we present a multi-interest GCN(MI-GCN) model for a recommendation, and it conducts high-order graph convolution operations in three sets of subgraphs. Users with similar interests and the corresponding interaction items belong to the identical subgraph. As for the formation of the subgraph, we adopt two varied clustering methods and the user feature to form a subgraph generation mechanism. This mechanism can generate three groups of differential subgraphs to divide users into multi-interest groups and make subgraph division more reasonable. We carry out massive experiments on three real-world datasets, demonstrating the effectiveness of our model. Experimental results confirm that our presented MI-GCN outperforms the state-of-the-art GCN-based recommendation models.

Cover Caption: The cover image is based on the Research Article Active forgetting via influence estimation for neural networks by Xianjia Meng et al., https://doi.org/10.1002/int.22981.

The rapidly exploding of user data, especially applications of neural networks, involves analyzing data collected from individuals, which brings convenience to life. Meanwhile, privacy leakage in the applications as a potential threat needs to be addressed urgently. However, removing private information from models is difficult once the user's sensitive data enters machine learning models, particularly neural networks. Most of the previous amnestic methods based on retraining require full access to the training set of the target model and have limited improvements in computational resources and time improvement. In this paper, we propose Scrubber, which removes sensitive data from the original model via influence estimation to produce an unlearning model that is approximately indistinguishable from the retrained model. S crubber builds on the essential concept of influence function and reformulates the influence estimation as a closed-form update of forgetting. For learned models with strictly convex loss functions, our approach theoretically guarantees the effectiveness of forgetting while empirically demonstrating forgetting performance. For models with non-convex losses, we relax strictly convex assumptions by applying a damping term that allows us to make approximate estimates with negligible errors from the original assumption. Furthermore, experiments show that S crubber only causes less than 1% and 3% accuracy drop with more than 80% forgetting rate on average for logistic regression models and convolutional neural networks. The accuracy drop is reduced by 2%–3% compared to most state-of-the-art methods.

To obtain reliable results from conflicting data in mobile crowdsensing, numerous truth discovery protocols have been proposed in the past decade. However, most of them do not consider the data privacy of entities involved (e.g., workers and servers), and several existing privacy-preserving truth discovery protocols either provide limited privacy protection or have heavy computation and communication overheads due to iterative computation and transmission over large ciphertexts. In this paper, we aim to propose privacy-preserving and lightweight truth discovery protocols to tackle the above problems. Specifically, we carefully design an anonymization protocol named AnonymTD to delink workers from their data, where workers’ data are computed and transmitted without complicated encryption. To further reduce each worker's overheads in the scenarios where workers are willing to share their weights, we resort to the perturbation technology to propose a more lightweight truth discovery protocol named PerturbTD. Based on workers’ perturbed data, two cloud servers in PerturbTD complete most of the workload of truth discovery together, which avoids the frequent involvement of workers. The theoretical analysis and the comparative experiments in this paper demonstrate that our two protocols can achieve our security goals with low computation and communication overheads.

Dynamic Searchable Symmetric Encryption (DSSE) has gained increasing popularity as it enables users to perform both file updates and ciphertext retrieval over encrypted data . However, existing DSSE schemes still lead to privacy leakage ( e.g., forward and backward privacy) in the dynamic setting. Some forward and backward secure DSSE schemes have been proposed, but still cannot support the keyword shielding flexibly. To solve this challenging issue, we propose a Forward and Backward Authorized Keyword Search (FB-AKS) scheme with recoverable keyword shielding by using trapdoor permutations and puncturable encryption in this paper. Compared with existing forward and backward private schemes, FB-AKS achieves keyword authorization flexibly ( e.g., keyword shielding, keyword un-shielding). The formal security analysis proves that FB-AKS achieves forward and backward security. And extensive experiments demonstrate that FB-AKS has less computation and storage overheads .