Privacy-preserving Algorithms Research Articles

Using Privacy-Preserving Algorithms and Blockchain Tokens to Monetize Industrial Data in Digital Marketplaces

Using Privacy-Preserving Algorithms and Blockchain Tokens to Monetize Industrial Data in Digital Marketplaces

Privacy-Preserving Algorithms Can Facilitate Surgical Video Databases for Quality Improvement, Education, and Research.

Privacy-Preserving Algorithms Can Facilitate Surgical Video Databases for Quality Improvement, Education, and Research.

Frequency Oracle for Sensitive Data Monitoring (Student Abstract)

As data privacy issues grow, finding the best privacy preservation algorithm for each situation is increasingly essential. This research has focused on understanding the frequency oracles (FO) privacy preservation algorithms. FO conduct the frequency estimation of any value in the domain. The aim is to explore how each can be best used and recommend which one to use with which data type. We experimented with different data scenarios and federated learning settings. Results showed clear guidance on when to use a specific algorithm.

Efficient approximation and privacy preservation algorithms for real time online evolving data streams

Efficient approximation and privacy preservation algorithms for real time online evolving data streams

Differential privacy scheme using Laplace mechanism and statistical method computation in deep neural network for privacy preservation

Mountainous amounts of information are now available in hospitals, finance, counter-terrorism, education and many other sectors. Those information can offer a rich source for analysis and decision making. Such information contains user's sensitive and personal data as well. This emanates direct conflict with the user's privacy. Individual's privacy is their right. The existing privacy preserving algorithms works mainly on the numerical data and doesn't care about the categorical data. In addition, there is a heavy trade-off between privacy preservation and data utility. To overcome these issues, a deep neural network - statistical differential privacy (DNN−SDP) algorithm is proposed as the solution to disguise the individual's private and sensitive data. Both the numerical and categorical based human-specific data are considered and fed to the input layer of the neural network. The statistical methods weight of evidence and information value is applied in the hidden layer along with the random weight (wi) to get the initial perturbed data. This initially perturbed data is taken by Laplace computation based differential privacy mechanism as the input and provides the final perturbed data. Census income, bank marketing and heart disease datasets are used for experimentation. While comparing with the state-of-the-art methods, DNN−SDP algorithm provides 97.4% of accuracy with 98.2% of precision, 99% of recall rate and 98.7% of F-measure value. In addition, the fall-out rate, miss rate and false omission rate of the proposed algorithm are less than 4.1%. The DNN−SDP algorithm guarantees the privacy preservation along with data utility.

Research on Location Privacy Protection Technology in Wireless Sensor Networks Based on Big Data

The digital footprint in wireless sensor networks can bring great academic and commercial value, but it will also bring the risk of privacy disclosure to users. This study discusses the location privacy protection methods in location based service under multiple scenarios. In the experiment, a false location filtering algorithm for real-time location request scenarios and a false path generation model for offline location release scenarios are proposed. The false position filtering algorithm is implemented based on the idea of a large top heap selection query. The algorithm can construct false position candidate sets and filter false positions. The false track generation model combines the false position technology and the generative adversarial networks model, which mainly protects the user’s track data by synthesizing tracks. In the attacker’s recognition experiment of a real location. The minimum distance between the false locations generated by the algorithm proposed in the study is above 400 m and the generation time does not exceed 5 ms, generating a better set of false locations in terms of both effectiveness and efficiency. Compared with several commonly used privacy-preserving algorithms, the proposed algorithm has the lowest probability of being identified with real locations, with no more than 21% overall, and is almost independent of the k value. the recognition accuracy of the trajectory user link task decreases from over 90% to about 34%, indicating that the proposed fake trajectory generation model can effectively protect users’ data privacy. The experimental results demonstrate that the algorithm and model proposed in the study can quickly generate physically dispersed and semantically diverse sets of fake locations and effectively protect users’ trajectory privacy, which is important for users’ digital footprint privacy protection.

Preserve data-while-sharing: An Efficient Technique for Privacy Preserving in OSNs

Online Social Networks (OSNs) have become one of the major platforms for social interactions, such as building up relationships, sharing personal experiences, and providing other services. Rapid growth in Social Network has attracted various groups like the scientific community and business enterprise to use these huge social network data to serve their various purposes. The process of disseminating extensive datasets from online social networks for the purpose of conducting diverse trend analyses gives rise to apprehensions regarding privacy, owing to the disclosure of personal information disclosed on these platforms. Privacy control features have been implemented in widely used online social networks (OSNs) to empower users in regulating access to their personal information. Even if Online Social Network owners allow their users to set customizable privacy, attackers can still find out users’ private information by finding the relationships between public and private information with some background knowledge and this is termed as inference attack. In order to defend against these inference attacks this research work could completely anonymize the user identity.
 This research work designs an optimization algorithm that aims to strike a balance between self-disclosure utility and their privacy. This research work proposes two privacy preserving algorithms to defend against an inference attack. The research work design an Privacy-Preserving Algorithm (PPA) algorithm which helps to achieve high utility by allowing users to share their data with utmost privacy. Another algorithm-Multi-dimensional Knapsack based Relation Disclosure Algorithm (mdKP-RDA) that deals with social relation disclosure problems with low computational complexity. The proposed work is evaluated to test the effectiveness on datasets taken from actual social networks. According on the experimental results, the proposed methods outperform the current methods.

PPVC: Towards a Personalized Local Differential Privacy-Preserving Scheme for V2G Charging Networks

The rapid development of electric vehicles provides users with convenience of life. When users enjoy the V2G charging service, privacy leakage of their charging location is a crucial security issue. Existing privacy-preserving algorithms for EV access to charging locations suffer from the problem of nondefendable background knowledge attacks and privacy attacks by untrustworthy third parties. We propose a personalized location privacy protection scheme (PPVC) based on differential privacy to meet users’ personalized EV charging requirements while protecting their privacy. First, by constructing a decision matrix, PPVC describes recommended routes’ utility and privacy effects. Then, a utility model is constructed based on the multiattribute theory. The user’s privacy preferences are integrated into the model to provide the route with the best utility. Finally, considering the privacy preference needs of users, the Euclidean distance share is used to assign appropriate privacy budgets to users and determine the generation range of false locations to generate the service request location with the highest utility. The experimental results show that the proposed personalized location privacy protection scheme can meet the service demands of users while reasonably protecting their privacy to provide higher service quality. Compared with existing solutions, PPVC improves the charging efficiency by up to 25%, and 8% at the same privacy protection level.

A negative survey based privacy preservation method for topology of social networks

With the development of social platforms, the social network has aroused wide attention. Since social networks contain a lot of personal sensitive information, many privacy preservation methods have been designed for social networks to allay concerns about privacy disclosure of people. However, most of the existing methods disturb the social networks too much to ensure the utility of social networks. To this end, we propose a negative survey based privacy preservation method, called NetNS, to preserve the topology privacy of social networks, where a dedicated negative survey model is developed to disturb edges in social networks in order to preserve the topology privacy of them. The theoretical analysis indicates that the developed NetNS is efficient, and can resist two common graph structure attacks including friendship attack and subgraph attack, while empirical studies conducted on three real-world social networks show that compared to six existing privacy preservation algorithms tailored for the topology of social networks, the developed NetNS can provide disturbed social networks with better utility while achieving same privacy preservation level.

Ethics of artificial intelligence

The general objective of the research was to determine the advances related to the Ethics of Artificial Intelligence. The most powerful countries are investing large amounts of economic resources in the development of artificial intelligence. Methodology, in this research, 49 documents have been selected, carried out in the period 2018 - 2023; including: scientific articles, review articles and information from websites of recognized organizations. Results, the ethics of artificial intelligence is supported by various countries. Furthermore, ChatGPT is considered a major threat in the automation of academic document preparation. Conclusions, about the general objective of the research is to determine the advances related to the Ethics of Artificial Intelligence, it is important to consider the transparency and risks of applying AI. In addition, consider ethical aspects such as the Recommendation on the Ethics of Artificial Intelligence was adopted by UNESCO's General Conference at its 41st session. The European Union (EU) is considering a new legal framework about regulations on the development and use of artificial intelligence. ChatGPT is an AI tool that needs to be carefully evaluated for its impact on education and other human activities. About the first specific objective of the research was to identify the countries that invest the most money in artificial intelligence, there are Japan, Singapore, China, India, Russia, Australia, Unite States of America, and the European Union. About the second specific objective of the research was to determine the risks and requirements of artificial intelligence, the risks are black-box models, privacy violations, bias and discrimination and the requirements are algorithmic transparency, human understandable explanations, privacy-preserving algorithms, data cooperatives, algorithmic fairness.

Review on security of federated learning and its application in healthcare

Artificial intelligence (AI) has led to a high rate of development in healthcare, and good progress has been made on many complex medical problems. However, there is a lack of patient electronic medical records standards and legal and ethical requirements for patient information privacy. The fragmentation of medical data has hindered the widespread use of AI in healthcare. Federated learning has emerged to address the problem of data fragmentation, and in combination with privacy-preserving algorithms can go a long way to address privacy concerns. Federated learning can also be combined with blockchain, edge computing, and other technologies to improve security and computational efficiency. The paper combs through recent papers on federated learning and medical industry applications, outline several architectures and classification models of federated learning, researches the application of federated learning in healthcare, and analyses the advantages of federated learning in medical applications. The paper also analyzes the security of federated learning and medical applications and analyses the various risks and attacks faced by the applications. Several standard privacy protection methods are introduced, and the current stage of the application of federated learning and healthcare is discussed. Finally, a summary and outlook on the application and security of federated learning in healthcare are presented.

Efficient privacy preserving algorithms for hiding sensitive high utility itemsets

Utility-driven mining is a powerful data mining technique that aims to extract significant and valuable knowledge from different kinds of datasets. Nevertheless, analyzing datasets with sensitive or private information may raise security and privacy concerns. To balance utility maximization and privacy preservation, Privacy Preserving Utility Mining (PPUM) has been presented. The primary objective of PPUM algorithms is to conceal the sensitive knowledge that can be found via the application of utility mining algorithms to sensitive data. However, the current PPUM literature shows a paucity of privacy preserving algorithms scalable and efficient enough to handle large and dense datasets. Based on this perspective, this paper proposes three heuristic-based algorithms, namely Selecting the Most Real item sensitive utility First (SMRF), Selecting the Least Real item sensitive utility First (SLRF), and Selecting the most Desirable Item First (SDIF), to efficiently conceal all Sensitive High utility Itemsets (SHIs) while mitigating the projected detrimental impact on the non-sensitive information. The proposed algorithms rely on a novel concept, called Real Item Sensitive Utility (RISU), to effectively select a definite victim item for each SHI during the whole sanitization process. Furthermore, a new sensitive dataset sorting technique is proposed to reduce the time needed to find suitable transactions for sanitization. Through comprehensive experimental evaluations with state-of-the-arts, the viability of the proposed three algorithms was testified. The acquired findings clearly demonstrate the efficacy of the proposed algorithms in terms of reducing the sanitization time and side effects.

Privacy protection in intelligent vehicle networking: A novel federated learning algorithm based on information fusion

Federated learning is an effective technique to solve the problem of information fusion and information sharing in intelligent vehicle networking. However, most of the existing federated learning algorithms generally have the risk of privacy leakage. To address this security risk, this paper proposes a novel personalized federated learning with privacy preservation (PDP-PFL) algorithm based on information fusion. In the first stage of its execution, the new algorithm achieves personalized privacy protection by grading users’ privacy based on their privacy preferences and adding noise that satisfies their privacy preferences. In the second stage of its execution, PDP-PFL performs collaborative training of deep models among different in-vehicle terminals for personalized learning, using a lightweight dynamic convolutional network architecture without sharing the local data of each terminal. Instead of sharing all the parameters of the model as in standard federated learning, PDP-PFL keeps the last layer local, thus adding another layer of data confidentiality and making it difficult for the adversary to infer the image of the target vehicle terminal. It trains a personalized model for each vehicle terminal by “local fine-tuning”. Based on experiments, it is shown that the accuracy of the proposed new algorithm for PDP-PFL calculation can be comparable to or better than that of the FedAvg algorithm and the FedBN algorithm, while further enhancing the protection of data privacy.

Quantum machine learning with differential privacy

Quantum machine learning (QML) can complement the growing trend of using learned models for a myriad of classification tasks, from image recognition to natural speech processing. There exists the potential for a quantum advantage due to the intractability of quantum operations on a classical computer. Many datasets used in machine learning are crowd sourced or contain some private information, but to the best of our knowledge, no current QML models are equipped with privacy-preserving features. This raises concerns as it is paramount that models do not expose sensitive information. Thus, privacy-preserving algorithms need to be implemented with QML. One solution is to make the machine learning algorithm differentially private, meaning the effect of a single data point on the training dataset is minimized. Differentially private machine learning models have been investigated, but differential privacy has not been thoroughly studied in the context of QML. In this study, we develop a hybrid quantum-classical model that is trained to preserve privacy using differentially private optimization algorithm. This marks the first proof-of-principle demonstration of privacy-preserving QML. The experiments demonstrate that differentially private QML can protect user-sensitive information without signficiantly diminishing model accuracy. Although the quantum model is simulated and tested on a classical computer, it demonstrates potential to be efficiently implemented on near-term quantum devices [noisy intermediate-scale quantum (NISQ)]. The approach’s success is illustrated via the classification of spatially classed two-dimensional datasets and a binary MNIST classification. This implementation of privacy-preserving QML will ensure confidentiality and accurate learning on NISQ technology.

Framework of Geofence Service using Dummy Location Privacy Preservation in Vehicular Cloud Network

With the increasing prevalence of different mobile apps, many applications require users to enable the location service on their devices. For example, the geofence service can be defined as establishing virtual geographical boundaries. Enabling this service triggers entering and exiting the boundary area and notifies the users and trusted third parties. The foremost concern of using geofence is the privacy of location coordinates shared among different applications. In this paper, a framework called ‘TIET-GEO’ is proposed that allows users to define the geofence boundary; in addition, it monitors Global Positioning System (GPS) devices in real-time when they enter/exit a specific area. The proposed framework also proposes a dummy privacy preservation algorithm to generate K-dummy locations around the real trajectories when the user requests the Point Of Interest (POI) from the Location-Based Services (LBS). This article aims to enhance the location privacy preservation in geofence service, by generating a k-dummy location around the user location based on the radius size of the geofence area. The proposed framework uses token keys authentication to authorize the users in the Vehicular Cloud Network (VCN) service by generating secret token keys authentication between the client and services. The results obtained show the effectiveness of the proposed framework was on parameters like flexibility and reliability of responses from different sources, such as smart IoT devices and datasets.

DPBA-WGAN: A Vector-Valued Differential Private Bilateral Alternative Scheme on WGAN for Image Generation

The large amount of sensitive personal information used in deep learning models has attracted considerable attention for privacy security. Sensitive data may be memorialized or encoded into the parameters or generation of the Wasserstein Generative Adversarial Networks (WGAN), which can be prevented by implementing privacy-preserving algorithms during the parameter training process. Meanwhile, the model is also expected to obtain effective generated results. We propose a vector-valued differential private bilateral alternative (DPBA) algorithm, a novel perturbation method for the training process. The vector-valued Gaussian (VVG) noise involving functional structure information is injected into the WGAN to generate data with privacy protection, and the model is verified to satisfy differential privacy. The bilateral alternative noise can eventually randomly perturb the gradient and generates informative feature-rich samples. The dynamic noise and vector-based perturbation approach ensures privacy strength. After extensive evaluation, our algorithm outperformed state-of-the-art techniques in terms of usability metrics for all validation datasets. The downstream classification accuracy for the generated Modified National Institute of Standards and Technology (Mnist) was 97.04%, whereas that for the Fashion Modified National Institute of Standards and Technology (Fashion-Mnist) dataset was 80.91%. Mnist improved the average accuracy of the neural network classifier by at least 16.81%, and Fashion-Mnist by at least 3.55%. In the multichannel generation tasks, the binary classification accuracy improved by at least 10.4% compared to CelebFaces Attributes (CelebA), and the accuracy of the Street View House Numbers (SVHN) was as high as 86.1%. The perturbation method proved highly resilient to gradient attack recovery under simulated gradient attacks.

Multi-Robot Privacy-Preserving Algorithms Based on Federated Learning: A Review

Multi-Robot Privacy-Preserving Algorithms Based on Federated Learning: A Review

A privacy preservation framework for feedforward-designed convolutional neural networks

A feedforward-designed convolutional neural network (FF-CNN) is an interpretable neural network with low training complexity. Unlike a neural network trained using backpropagation (BP) algorithms and optimizers (e.g., stochastic gradient descent (SGD) and Adam), a FF-CNN obtains the model parameters in one feed-forward calculation based on two methods of data statistics: subspace approximation with adjusted bias and least squares regression. Currently, models based on FF-CNN training methods have achieved outstanding performance in the fields of image classification and point cloud data processing. In this study, we analyze and verify that there is a risk of user privacy leakage during the training process of FF-CNN and existing privacy-preserving methods for model gradients or loss functions do not apply to FF-CNN models. Therefore, we propose a securely forward-designed convolutional neural network algorithm (SFF-CNN) to protect the privacy and security of data providers for the FF-CNN model. Firstly, we propose the DPSaab algorithm to add the corresponding noise to the one-stage Saab transform in the FF-CNN design for improved protection performance. Secondly, because noise addition brings the risk of model over-fitting and further increases the possibility of privacy leakage, we propose the SJS algorithm to filter the input features of the fully connected model layer. Finally, we theoretically prove that the proposed algorithm satisfies differential privacy and experimentally demonstrate that the proposed algorithm has strong privacy protection. The proposed algorithm outperforms the compared deep learning privacy-preserving algorithms in terms of utility and robustness.

Federated Learning in Medical Imaging: Part I: Toward Multicentral Health Care Ecosystems.

With recent developments in medical imaging facilities, extensive medical imaging data are produced every day. This increasing amount of data provides an opportunity for researchers to develop data-driven methods and deliver better health care. However, data-driven models require a large amount of data to be adequately trained. Furthermore, there is always a limited amount of data available in each data center. Hence, deep learning models trained on local data centers might not reach their total performance capacity. One solution could be to accumulate all data from different centers into one center. However, data privacy regulations do not allow medical institutions to easily combine their data, and this becomes increasingly difficult when institutions from multiple countries are involved. Another solution is to use privacy-preserving algorithms, which can make use of all the data available in multiple centers while keeping the sensitive data private. Federated learning (FL) is such a mechanism that enables deploying large-scale machine learning models trained on different data centers without sharing sensitive data. In FL, instead of transferring data, a general model is trained on local data sets and transferred between data centers. FL has been identified as a promising field of research, with extensive possible uses in medical research and practice. This article introduces FL, with a comprehensive look into its concepts and recent research trends in medical imaging.

DESIGN AND DEVELOPMENT OF ANTI-LOCK BRAKING SYSTEM FOR ELECTRIC VEHICLE

Abstract—Recently, privacy-preserving association rules mining algorithms have been proposed to support data privacy. However, the algorithms have an additional overhead to insert fake items (or fake transactions) and cannot hide data frequency. In this paper, we propose a privacy-preserving association rule mining algorithm for encrypted data in cloud computing. For association rule mining, we utilize Apriori algorithm by using the Elgamal cryptosystem, without additional fake transactions. Thus the proposed algorithm can guarantee both data privacy and query privacy, while concealing data frequency. We show that the proposed algorithm achieves about 3-5 times better performance than the existing algorithm, in terms of association rule mining time. Keywords- association rule mining; Apriori algorithm; encrypted data; cloud computing; Elgamal cryptosystem;