Privacy-preserving Model Research Articles

Two Level Privacy Preserving Model for Association Rule Mining in Cloud Using Deep-SVM Method

Two Level Privacy Preserving Model for Association Rule Mining in Cloud Using Deep-SVM Method

PRIMϵ: Novel Privacy-Preservation Model With Pattern Mining and Genetic Algorithm

PRIMϵ: Novel Privacy-Preservation Model With Pattern Mining and Genetic Algorithm

LDCML: A Novel AI-Driven Approach form Privacy-Preserving Anonymization of Quasi-Identifiers

Introduction: the exponential growth of data generation has led to an escalating concern for data privacy on a global scale. This work introduces a pioneering approach to address the often overlooked data privacy leakages associated with quasi-identifiers, leveraging artificial intelligence, machine learning and data correlation analysis as foundational tools. Traditional data privacy measures predominantly focus on anonymizing sensitive attributes and exact identifiers, leaving quasi-identifiers in their raw form, potentially exposing privacy vulnerabilities.Objective: the primary objective of the presented work, is to anonymise the quasi-identifiers to enhance the overall data privacy preservation with minimal data utility degradation.Methods: In this study, the authors propose the integration of ℓ-diversity data privacy algorithms with the OPTICS clustering technique and data correlation analysis to anonymize the quasi-identifiers.Results: to assess its efficacy, the proposed approach is rigorously compared against benchmark algorithms. The datasets used are - Adult dataset and Heart Disease Dataset from the UCI machine learning repository. The comparative metrics are - Relative Distance, Information Loss, KL Divergence and Execution Time.Conclusion: the comparative performance evaluation of the proposed methodology demonstrates its superiority over established benchmark techniques, positioning it as a promising solution for the requisite data privacy-preserving model. Moreover, this analysis underscores the imperative of integrating artificial intelligence (AI) methodologies into data privacy paradigms, emphasizing the necessity of such approaches in contemporary research and application domains

Learning from vertically distributed data across multiple sites: An efficient privacy-preserving algorithm for Cox proportional hazards model with variable selection

ObjectiveTo develop a lossless distributed algorithm for regularized Cox proportional hazards model with variable selection to support federated learning for vertically distributed data. MethodsWe propose a novel distributed algorithm for fitting regularized Cox proportional hazards model when data sharing among different data providers is restricted. Based on cyclical coordinate descent, the proposed algorithm computes intermediary statistics by each site and then exchanges them to update the model parameters in other sites without accessing individual patient-level data. We evaluate the performance of the proposed algorithm with (1) a simulation study and (2) a real-world data analysis predicting the risk of Alzheimer’s dementia from the Religious Orders Study and Rush Memory and Aging Project (ROSMAP). Moreover, we compared the performance of our method with existing privacy-preserving models. ResultsOur algorithm achieves privacy-preserving variable selection for time-to-event data in the vertically distributed setting, without degradation of accuracy compared with a centralized approach. Simulation demonstrates that our algorithm is highly efficient in analyzing high-dimensional datasets. Real-world data analysis reveals that our distributed Cox model yields higher accuracy in predicting the risk of Alzheimer’s dementia than the conventional Cox model built by each data provider without data sharing. Moreover, our algorithm is computationally more efficient compared with existing privacy-preserving Cox models with or without regularization term. ConclusionThe proposed algorithm is lossless, privacy-preserving and highly efficient to fit regularized Cox model for vertically distributed data. It provides a suitable and convenient approach for modeling time-to-event data in a distributed manner.

A unified privacy preserving model with AI at the edge for Human-in-the-Loop Cyber-Physical Systems

With the proliferation of personal Internet of Things (IoT) devices and their wide adoption among society, the original notion of IoT has been reshaped, giving rise to new IoT-based paradigms like Human-in-the-Loop Cyber-Physical Systems (HiTLCPS). While these systems can bring benefits and positively influence people, their pervasive nature raises significant privacy concerns, especially regarding the acquisition and processing of data. Although privacy-preserving mechanisms have been proposed for these implementations, the existing approaches tend to focus on either data acquisition or data processing. However, to date, no solution has encompassed the entire process. In this regard, this paper presents a unified privacy-preserving model for HiTLCPS. This approach integrates a human-centric mechanism to control and make data acquisition and sharing tasks transparent with a state inference process supported by Artificial Intelligence (AI) in the edge. A set of assessments were conducted in actual implementation to evaluate the feasibility of the model. Our findings reveal that our federated learning approach is a suitable solution compared to the traditional approach based on machine learning at the small cost of 3.27% of average accuracy. Finally, this paper concludes by providing a roadmap towards integrating HiTLCPS from diverse contexts, including a state inference process closer to the user domain.

Resource management in multi-heterogeneous cluster networks using intelligent intra-clustered federated learning

The heterogeneous cluster networks (HCN) have recently benefited from federated learning (FL). On distributed data, FL is used to train privacy-preserving models. In heterogeneous networks (HetNet) and the Internet of Things (IoT), FL implementation is challenged by resource optimization, robustness, and security issues. There is a significant risk of data security being compromised by disregarding the nodes’ clustering behavior and quickly varying asynchronous streaming data. Moreover, in HCN-based wireless sensor networks (WSNs), FL enhances asynchronous node performance. Using naturally clustered HCN, distributed nodes train a local and global model collectively. In this paper, we propose a Intra-Clustered FL (ICFL) model. By optimizing computation and communication, ICFL selects heterogeneous FL nodes in each cluster. Despite heterogeneous data, it is highly robust. There are currently no FL frameworks that can handle varying data quality across devices and non-identical distributions. With ICFL, sensitive asynchronous data is not exposed to possible misuse while adapting to changing environments. In addition to being time-efficient, our strategy requires low-power computing nodes. According to our extensive simulation results, ICFL performs better than FedCH in terms of computational performance and provides flexible conditions under which ICFL is more efficient in terms of communication. In extensive testing, ICFL decreased training rounds by 62% and increased accuracy by 6.5%. It can execute evaluations 7.46 times faster than existing models, and its average accuracy has increased by 4.39%. A resource-aware FL system can be successfully implemented in real-time applications according to our research.

E-DPNCT: an enhanced attack resilient differential privacy model for smart grids using split noise cancellation

High frequency reporting of energy consumption data in smart grids can be used to infer sensitive information regarding the consumer’s life style and poses serious security and privacy threats. Differential privacy (DP) based privacy models for smart grids ensure privacy when analysing energy consumption data for billing and load monitoring. However, DP models for smart grids are vulnerable to collusion attack where an adversary colludes with malicious smart meters and un-trusted aggregator in order to get private information from other smart meters. We first show the vulnerability of DP based privacy model for smart grids against collusion attacks to establish the need of a collusion resistant privacy model. Then, we propose an Enhanced Differential Private Noise Cancellation Model for Load Monitoring and Billing for Smart Meters (E-DPNCT) which not only provides resistance against collusion attacks but also protects the privacy of the smart grid data while providing accurate billing and load monitoring. We use differential privacy with a split noise cancellation protocol with multiple master smart meters (MSMs) to achieve collusion resistance. We propose an Enhanced Differential Private Noise Cancellation Model for Load Monitoring and Billing for Smart Meters (E-DPNCT) to protect the privacy of the smart grid data using a split noise cancellation protocol with multiple master smart meters (MSMs) to provide accurate billing and load monitoring and resistance against collusion attacks. We did extensive comparison of our E-DPNCT model with state of the art attack resistant privacy preserving models such as EPIC for collusion attack. We simulate our E-DPNCT model with real time data which shows significant improvement in privacy attack scenarios. Further, we analyze the impact of selecting different sensitivity parameters for calibrating DP noise over the privacy of customer electricity profile and accuracy of electricity data aggregation such as load monitoring and billing.

Differential privacy may have a potential optimization effect on some swarm intelligence algorithms besides privacy-preserving

Differential privacy (DP), as a promising privacy-preserving model, has attracted great interest from researchers in recent years. At present, research on the combination of deep learning and DP is active. In contrast, another widely used artificial intelligence technique, swarm intelligence (SI), has received little attention in the context of DP even though it also triggers privacy concerns. For this reason, this paper attempts to combine DP and SI for the first time and proposes a general differentially private swarm intelligence algorithm framework (DPSIAF). By utilizing the exponential mechanism, this framework can easily develop existing SI algorithms into the private versions. As examples, we apply it to four popular SI algorithms, and the related analyses demonstrate the effectiveness of our DPSIAF. More importantly, our private algorithms also exhibit some interesting experimental phenomena. In some cases, their performance is not strictly affected by the privacy budget, and one of them even outperforms its non-private version. These findings are different from the conventional cognition, which indicates DP owns the potential to serve as an optimization tool for some SI algorithms. Our study may provide a new perspective on DP and promote the synergy between metaheuristic optimization community and privacy computing community.

Multi-level personalized k-anonymity privacy-preserving model based on sequential three-way decisions

K-anonymity is a widely used privacy-preserving technique which defends against linking attacks by suppression and generalization. The existing k-anonymity algorithms prevent attackers from illegally obtaining private information by constraining at least k records in an equivalence group. However, this unified anonymity method ignores individual differences and leads to a large amount of information loss. To this end, we introduce sequential three-way decisions into k-anonymity, using a dynamic k-value sequence instead of the fixed k-value to achieve personalized k-anonymity. Specifically, we first construct a hierarchical decision table for k-anonymity by attribute generalization trees and sensitive decision values provided with users. Then, we propose a multi-level personalized k-anonymity privacy-preserving model based on sequential three-way decisions, where we anonymize the partitioning granular data with a dynamic k-value sequence, respectively. Furthermore, we present three practical algorithms to implement the proposed model and discuss the differences between them. Finally, the experimental results demonstrate that the proposed model not only provides a more flexible anonymization method to achieve personalized anonymity, but greatly reduces the information loss. This study provides a complete framework for multi-level privacy protection and enriches the application of sequential three-way decisions.

A Federated Learning Approach With Imperfect Labels in LoRa-Based Transportation Systems

Intelligent Transportation System (ITS) helps to improve vehicle health, driver safety, and passenger comfort. Remotely sharing the information of ITS to train the machine and deep learning models hamper data privacy and generate security threats to the passenger, driver, and vehicle owners. Moreover, sharing the information requires huge networking resources such as high data rate, low latency, and low packet loss. Federated learning provides privacy-preserving model training on the vehicle without sharing the information. However, due to poor annotation mechanisms, federated learning may suffer from imperfect labels. This paper proposes a federated learning approach for ITS that can handle imperfect labels in the datasets of the participants. The approach also uses a Long-Range network to provide communication efficient connectivity. The approach initially estimates class-wise centroids of the datasets at the participants and server and then identifies participants with imperfect labels using similarity scores. Such participants demand the fraction of the correctly annotated dataset at the server to improve performance. We further derive the expression for the optimal fraction of the dataset requested by a participant. We finally verify the effectiveness of the proposed approach using the existing model and publicly available dataset.

MistNet: A superior edge-cloud privacy-preserving training framework with one-shot communication

Classical federated learning methods aggregate decentralized data from different devices into a central location for efficient training. However, these approaches raise significant concerns regarding data privacy and requires extensive communication between edge devices and the cloud, leading to high communication costs. To conquer these challenges, we propose an edge-cloud privacy-preserving model training system that enables the cloud and edge devices to collaboratively perform deep neural network(DNN) training, namely MistNet, which has the following advantages. (1) Different from previous works that required frequent communication between the edge and cloud, MistNet requires only one-shot communication. (2) MistNet introduces the rigorous local differential privacy (LDP) technique to the intermediate feature maps, guaranteeing that the user’s local data and model parameters remain undisclosed. (3) In MistNet, the feature extractor is transferred from pre-trained models that are designed for similar application domains. This feature extractor remains fixed during training, thereby eliminating the requirement to synchronize feature extractors across different devices. Furthermore, we first design an object detection algorithm based on Yolov5 with the MistNet framework. Finally, the extensive experiments results on multiple models and datasets demonstrate that by choosing an appropriate partition layer and privacy budget, MistNet achieves lower communication and faster convergence, as well as acceptable model utility while greatly reducing privacy leakage from the released intermediate features. Our framework is primarily designed for image data and has already been applied to Plato. You can find the source code on GitHub page at https://github.com/TL-System/plato.

Privacy-preserving model for biometric-based authentication and Key Derivation Function

Bio-cryptosystems often save the biometric template for authentication and generally employ randomly generated keys to encrypt and sign data. This method raises privacy protection concerns. Furthermore, for a system secured by the usage of a cryptographic key, losing the key often has disastrous consequences. To overcome the privacy issues, and allow the secure recovery of lost keys, we design a Key Derivation Function to extract a key from biometric data: a new method – based on clustering algorithms – detects consistent and discriminative features from biometric characteristics to create a code. Then, HMAC-SHA256 (as specified by the National Institute of Standards and Technology) generates a standard key from the code. To reproduce the code at future times (that also serves for authentication), the Key Derivation Function stores helper data with the guarantee of privacy. Indeed, with a private face dataset, the probability of generating the code with only the helper data is less than 2−300, and less than 2−246 for a subset of the YouTube Face database. Moreover, on the private database and the tested users from the YouTube Face database, our system has a false acceptance rate of 0%. It corrects up to 40.3% of noise levels on the private database and has good management of the inter-user variability.

Harnessing the power of federated learning to advance technology

Federated Learning (FL) has emerged as a transformative paradigm in machine learning, advocating for decentralized, privacy-preserving model training. This study provides a comprehensive evaluation of contemporary FL frameworks TensorFlow Federated (TFF), PySyft, and FedJAX across three diverse datasets: CIFAR-10, IMDb reviews, and the UCI Heart Disease dataset. Our results demonstrate TFF's superior performance on image classification tasks, while PySyft excels in both efficiency and privacy for textual data. The study underscores the potential of FL in ensuring data privacy and model performance, yet emphasizes areas warranting improvement. As the volume of edge devices escalates and the need for data privacy intensifies, refining and expanding FL frameworks become essential for future machine learning deployments.

Privacy-preserving DeepFake face image detection

All the existing models for DeepFake detection focus on plaintext faces. However, outsourced computing is usually considered in practical applications for DeepFake detection and the input data may contain private and sensitive information. Thus, a privacy-preserving model named Secure DeepFake Detection Network (SecDFDNet) is proposed for the first time in this paper. The SecDFDNet uses the additive secret sharing method for secure DeepFake face detection. Specifically, firstly, some multi-party secure interaction protocols are designed for non-linear activation functions, i.e., SecReLU for ReLU function, SecSigm for sigmoid function, SecSpatial for spatial attention, and SecChannel for channel attention. Their security is proved in theory. Our protocols have low communication and space complexity. Then, the SecDFDNet model is proposed by using the designed secure protocols and trained plaintext DeepFake detection network (DFDNet). The experimental results show that the proposed SecDFDNet can detect DeepFake faces without revealing anything of private input, achieve the same accuracies as the plaintext DFDNet and outperform some existing models. The source code is available at https://github.com/imagecbj/Privacy-Preserving-DeepFake-Face-Image-Detection.

A Novel Blockchain-Assisted Aggregation Scheme for Federated Learning in IoT Networks

With the wide range of Internet of things (IoT) applications, Federated Learning (FL) is commonly adopted to protect the privacy of IoT data. FL enables privacy-preserving model training while keeping the data locally available. To alleviate the additional load caused by FL, an improved hierarchical aggregation framework is presented in this paper to decentralize the model aggregation tasks based on end-device clusters. However, when applying FL to IoT networks, how to keep high efficiency and reliability remains open challenges due to a large number and vulnerability of IoT end-devices. In this paper, we propose a blockchain-assisted aggregation scheme for FL in IoT networks, where the aggregation node selection is applied for efficiency improvement as well as blockchain for performance verification. During model aggregation, a selection strategy is obtained by the Deep Deterministic Policy Gradient (DDPG) algorithm and aims to select the optimal subset of IoT end-devices based on multiple metrics. Furthermore, a new performance verification based on the characteristics of blockchain is applied to achieve mutual verification among a number of untrustworthy nodes with the optimal stopping theory, which provides reliable model performance proofs. Simulation results show that the proposed scheme can maintain FL efficiency and reduce the system latency while protecting data privacy.

MLChain: a privacy-preserving model learning framework using blockchain

MLChain: a privacy-preserving model learning framework using blockchain

Towards Federated Learning and Multi-Access Edge Computing for Air Quality Monitoring: Literature Review and Assessment

Systems for monitoring air quality are essential for reducing the negative consequences of air pollution, but creating real-time systems encounters several challenges. The accuracy and effectiveness of these systems can be greatly improved by integrating federated learning and multi-access edge computing (MEC) technology. This paper critically reviews the state-of-the-art methodologies for federated learning and MEC-enabled air quality monitoring systems. It discusses the immense benefits of federated learning, including privacy-preserving model training, and MEC, such as reduced latency and improved response times, for air quality monitoring applications. Additionally, it highlights the challenges and requirements for developing and implementing real-time air quality monitoring systems, such as data quality, security, and privacy, as well as the need for interpretable and explainable AI-powered models. By leveraging such advanced techniques and technologies, air monitoring systems can overcome various challenges and deliver accurate, reliable, and timely air quality predictions. Moreover, this article provides an in-depth analysis and assessment of the state-of-the-art techniques and emphasizes the need for further research to develop more practical and affordable AI-powered decentralized systems with improved performance and data quality and security while ensuring the ethical and responsible use of the data to support informed decision making and promote sustainability.

Quasi‐identifier recognition with echo chamber optimization‐based anonymization for privacy preservation of cloud storage

SummaryThe publication of data raises several security concerns, implying that when a reputable company offers data to a third party, personal information is not needed to be revealed. The fundamental issue that led to identity leakage is the connection of quasi‐identifiers (QIDs); however, the majority of researchers overlook the identification of accurate QIDs. This study developed a privacy preservation and clustering‐based quasi‐identification model based on the Echo Chamber optimization with a Z‐mixture parameter for privacy‐preserving combined data publishing to maintain the privacy of the data. The Echo chamber optimization is used to conduct a clustering‐based quasi‐identification to find the significant attributes especially, in determining the dimension of the solutions as well as the precise QIDs using the re‐identification risk rate as the fitness function. The loss of information is greatly reduced in terms of the metrics such as normalized certainty penalty metric, average equivalent class size metric, and discernibility metric. The developed optimized clustering‐based algorithm with the privacy preservation model extensively minimizes the leakage of private information and the utilization of data is well‐maintained compared with other existing algorithms.

A data privacy-preserving model driven on differential approach in cloud environment

A data privacy-preserving model driven on differential approach in cloud environment

Towards Fair and Selectively Privacy-Preserving Models Using Negative Multi-Task Learning (Student Abstract)

Deep learning models have shown great performances in natural language processing tasks. While much attention has been paid to improvements in utility, privacy leakage and social bias are two major concerns arising in trained models. In order to tackle these problems, we protect individuals' sensitive information and mitigate gender bias simultaneously. First, we propose a selective privacy-preserving method that only obscures individuals' sensitive information. Then we propose a negative multi-task learning framework to mitigate the gender bias which contains a main task and a gender prediction task. We analyze two existing word embeddings and evaluate them on sentiment analysis and a medical text classification task. Our experimental results show that our negative multi-task learning framework can mitigate the gender bias while keeping models’ utility.