Privacy-preserving Model Research Articles

Differential and TriPhase Adaptive Learning-Based Privacy-Preserving Model for Medical Data in Cloud Environment

This letter proposes a novel secure data protection model for privacy preservation in the cloud environment by partitioning, sanitizing, and analyzing the data effectively to improve the model’s privacy. Several experiments and comparisons of the proposed model with existing works indicate that it protects data with high accuracy, precision, recall, and F1-score up to 87.03%, 84.87%, 87.03%, and 85.00%, for diverse datasets with a relative improvement up to 15.89%, 27.73%, 15.89%, and 21.44%, respectively.

Blockchain and homomorphic encryption based privacy-preserving model aggregation for medical images.

Medical healthcare centers are envisioned as a promising paradigm to handle the massive volume of data for COVID-19 patients using artificial intelligence (AI). Traditionally, AI techniques require centralized data collection and training models within a single organization. This practice can be considered a weakness as it leads to several privacy and security concerns related to raw data communication. To overcome this weakness and secure raw data communication, we propose a blockchain-based federated learning framework that provides a solution for collaborative data training. The proposed framework enables the coordination of multiple hospitals to train and share encrypted federated models while preserving data privacy. Blockchain ledger technology provides decentralization of federated learning models without relying on a central server. Moreover, the proposed homomorphic encryption scheme encrypts and decrypts the gradients of the model to preserve privacy. More precisely, the proposed framework: (i) train the local model by a novel capsule network for segmentation and classification of COVID-19 images, (ii) furthermore, we use the homomorphic encryption scheme to secure the local model that encrypts and decrypts the gradients, (iii) finally, the model is shared over a decentralized platform through the proposed blockchain-based federated learning algorithm. The integration of blockchain and federated learning leads to a new paradigm for medical image data sharing over the decentralized network. To validate our proposed model, we conducted comprehensive experiments and the results demonstrate the superior performance of the proposed scheme.

Eiffel: Efficient and Fair Scheduling in Adaptive Federated Learning

Emerging machine learning (ML) technologies, in combination with the increasing computational power of mobile devices, lead to the extensive adoption of ML-based applications. Different from conventional model training that needs to collect all the user data in centralized cloud servers, federated learning (FL) has recently drawn increasing research attention as it enables privacy-preserving model training. With FL, decentralized edge devices in participation, train their model copies locally over their siloed datasets, and periodically synchronize the model parameters. However, model training is computationally extensive which easily drains the battery of mobile devices. In addition, due to the uneven distribution of siloed datasets, the shared model may become biased. To address the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">efficiency</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">fairness</i> concerns in a resource-constrained federated learning setting, in this paper, we propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Eiffel</i> to judiciously select mobile devices to participate in the global model aggregation, and adaptively adjust the frequency of local and global model updates. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Eiffel</i> aims to make scheduling and coordination for the federated learning towards both resource efficiency and model fairness. We have conducted theoretical analysis of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Eiffel</i> from the perspectives of fairness and convergence. Extensive experiments with a wide variety of real-world datasets and models, both on a networked prototype system and in a larger-scale simulated environment, have demonstrated that while maintaining similar accuracy performance, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Eiffel</i> outperforms existing baselines with respect to reducing communication overhead by up to 6× for higher efficiency and improving the fairness metric by up to 57% compared to the state-of-the-art algorithms.

Pragmatic evaluation of privacy preservation security models targeted towards fog-based deployments

Fog layer sits between cloud layer and edge-layer and responsible for selection of edge-nodes to process cloud tasks. Fog devices manage routers, gateways and other scheduling components, which makes them highly vulnerable to security attacks. Attackers inject malicious packets fog-server, middleware or sensing layers which causes a wide variety of attacks. These attacks include node capturing, signal jamming, node outage, authorization, selective forwarding, data disclosure etc. To remove these attacks, wide variety of solutions are proposed by researchers, which include authorization, cryptography, error correction, firewall, broadcast authentication, selective disclosure etc. Moreover, these solutions vary with respect to privacy and security quality metrics, attack prevention capabilities and deployment quality of service (QoS). Thus, testing and deployment of these solutions is time consuming, requires additional manpower for performance validation. Hence fog deployments require larger time-to-market and are costly than their corresponding cloud deployments. In order to reduce the time for testing and validation of these resilience techniques, this text reviews various fog security &amp; privacy preservation models and discusses their nuances, advantages, limitations and future research scopes. Furthermore it also performs a detailed performance comparison between the reviewed models, which assists in selecting best possible approach for a given application scenario. This text also recommends various fusion based approaches that can be applied to existing security and privacy models in order to further improve their performance. These approaches include hybridization, selective augmentation and Q-learning based models that assist in improving efficiency of encryption, privacy preservation, while maintaining high QoS levels.

Design of privacy preserving model based on clustering involved anonymization along with feature selection

Healthcare services has become a hug transformation due to the number of disease emerging at presently. Accordingly, enormous amount of data is generated regarding on the patient data. Hence, preserving privacy of patient data is the major concern to meet privacy requirements and to prevent healthcare data from numerous privacy attacks. To protect the information from privacy attack, various studies were performed to safeguard the patient's data but, these methods have several issues when preserving privacy that increases proportionally as the data dimension increased in multiple approaches, which produces less data quality, more information loss and more execution time.Thus, selecting relevant attributes plays an important factor for improving the efficiency of any preserving algorithm. In order to overcome these concerns, in this paper proposed a privacy preserving model based on clustering involved anonymization along with feature selection approaches. The proposed model consist of two phases such as, feature selection and anonymization. In the initial phase, the relevant features are selected using Symmetrical Uncertainty (SU) and the redundant features present in the dataset are removed utilizing Spearman's Correlation Coefficient. In the second phase, the privacy preservation is performed using Utility Preserved Anonymization (UPA) algorithm. Also, the proposed algorithm reduces the dimensionality of data to ease the process of forming clusters for anonymization. The experimental analysis using real time datasets to verifies the effectiveness of the proposed method. The results show high sensitivity (up to 98.63%) and high accuracy (up to 98%), allowing us to claim efficient attribute selection for anonymization. Hence proved the proposed method reduces clustering's complexity by removing the irrelevant attributes effectively.

Quorum-based model learning on a blockchain hierarchical clinical research network using smart contracts

BackgroundCollaborative privacy-preserving modeling across several healthcare institutions allows for the construction of more generalizable predictive models while protecting patient privacy. ObjectiveWe aim at addressing the site availability issue on a hierarchical network by designing an immutable/transparent/source-verifiable quorum mechanism. MethodsWe developed an approach to combine a hierarchical learning algorithm, a novel Proof-of-Quorum (PoQ) consensus protocol, and a design of blockchain smart contracts. We constructed QuorumChain as an example and evaluated the scenarios of site-unavailability during the initialization and/or iteration phases of the modeling process on three healthcare/genomic datasets. ResultsWhen one or more sites would become unavailable, HierarchicalChain could not function, whereas QuorumChain improved predictive correctness significantly (the full Area Under the receiver operating characteristic Curve, or AUC, improved from 0.068 to 0.441, all with p-values < 0.001). ConclusionBy constructing a quorum to continue the modeling process, QuorumChain possesses the capability to tackle the situation of sites being unavailable. It inherits the capability of learning on network-of-networks, improves learning continuity, and provides data/software immutability, transparency, and provenance, which can be important in expediting clinical research.

NAVIGATOR: an Italian regional imaging biobank to promote precision medicine for oncologic patients

NAVIGATOR is an Italian regional project boosting precision medicine in oncology with the aim of making it more predictive, preventive, and personalised by advancing translational research based on quantitative imaging and integrative omics analyses. The project’s goal is to develop an open imaging biobank for the collection and preservation of a large amount of standardised imaging multimodal datasets, including computed tomography, magnetic resonance imaging, and positron emission tomography data, together with the corresponding patient-related and omics-related relevant information extracted from regional healthcare services using an adapted privacy-preserving model. The project is based on an open-source imaging biobank and an open-science oriented virtual research environment (VRE). Available integrative omics and multi-imaging data of three use cases (prostate cancer, rectal cancer, and gastric cancer) will be collected. All data confined in NAVIGATOR (i.e., standard and novel imaging biomarkers, non-imaging data, health agency data) will be used to create a digital patient model, to support the reliable prediction of the disease phenotype and risk stratification. The VRE that relies on a well-established infrastructure, called D4Science.org, will further provide a multiset infrastructure for processing the integrative omics data, extracting specific radiomic signatures, and for identification and testing of novel imaging biomarkers through big data analytics and artificial intelligence.

A node trust factor linked privacy preservation model in 5G networks with a multi-stage authentication model

There has been a dramatic expansion in the number and frequency of relationships between entities in mobile telecommunications networks over the last several years. For theseinteractions to be profitable, the parties involved must be able to count on one another. Since mobile networks require trust and reputation models to create feasible communications in 5 G and beyond networks, a collection of entities can create chains of operations between cross-operators with privacy and trustworthiness through mobile telecommunication network models. A major impediment to widespread communication beyond 5 G networks is the absence of automated, efficient, and scalable models for creating security and trust. Multitenancy and active infrastructure sharing will be major facilitators of new business models as 5 G networks expand. New security concerns are emerging due to these new opportunities, which is one of the primary obstacles for the mass implementation of 5 G networks. There are several ways in which attackers might use 5G-enabled situations to undertake lateral moves and disrupt 5 G services and infrastructure resources. Existing trust and security models cannot deal with the dynamic nature of the 5 G infrastructural threats or the multi-tenant security issues. Hence, we propose a Mobile Node Trust Factor Linked Privacy Preservation model for 5 G multi-stage authentication models in this work. A threat model for multi-stage authentication scenarios in an underlying 5 G network infrastructure is used to demonstrate how this model can be used. For secure end-to-end communication, authorization and strong authentication are necessary. A combination of anonymous authentication and authorization is needed to preserve the privacy of mobile devices which share personal information. The proposed model is compared with the traditional models and the proposed model results exhibit better performance.

APRO: Secret Key Based Data Encryption and Mutual Authentication with Privacy Preservation Model in IoT Using Autoregressive Poor and Rich Optimization

An essential security component for the Internet of Things (IoT) is device authentication. IoT devices are widely dispersed in public and open areas, making them vulnerable to attacks. To detect the security violations, an authentication with a privacy preservation model is developed in this research using devised Autoregressive Poor and Rich Optimization (APRO)-based secret key. Here, the secret key used to perform the encryption process is generated by the proposed optimization algorithm in such a way that the data encryption mechanism is carried out by considering the Box-Cox transformation. However, the authentication process is achieved between the server and the Authorization center (AC) through the generation of authentication and verification messages. The original data is retrieved back by employing the decryption process along with the usage of the retrieval key. The proposed method achieved a minimum computation time of 0.062 sec, higher encryption quality of 92.56%, and minimum memory usage of 2.86MB.

Fuzz-classification (p, l)-Angel: An enhanced hybrid artificial intelligence based fuzzy logic for multiple sensitive attributes against privacy breaches

The inability of traditional privacy-preserving models to protect multiple datasets based on sensitive attributes has prompted researchers to propose models such as SLOMS, SLAMSA, (p, k)-Angelization, and (p, l)-Angelization, but these were found to be insufficient in terms of robust privacy and performance. (p, l)-Angelization was successful against different privacy disclosures, but it was not efficient. To the best of our knowledge, no robust privacy model based on fuzzy logic has been proposed to protect the privacy of sensitive attributes with multiple records. In this paper, we suggest an improved version of (p, l)-Angelization based on a hybrid AI approach and privacy-preserving approach like Generalization. Fuzz-classification (p, l)-Angel uses artificial intelligence based fuzzy logic for classification, a high-dimensional segmentation technique for segmenting quasi-identifiers and multiple sensitive attributes. We demonstrate the feasibility of the proposed solution by modelling and analyzing privacy violations using High-Level Petri Nets. The results of the experiment demonstrate that the proposed approach produces better results in terms of efficiency and utility.

A differential approach and deep neural network based data privacy-preserving model in cloud environment

A differential approach and deep neural network based data privacy-preserving model in cloud environment

Differential privacy medical data publishing method based on attribute correlation

The advent of the era of big data promotes the further development of medicine, and data release is an important step in it. The existing medical data release methods mostly use the k-anonymity model as the basis for data protection. With the advancement of technology, anonymous models are progressively less resistant to consistency attacks and background knowledge attacks. In order to better protect the private information of patients, this paper makes two major contributions: (1) The method of calculating the correlation between attributes is used to ensure the validity of the data after the data is released; (2) On the basis of the previous step, combined with the difference privacy-preserving model and tree model, this paper proposes an attribute association-based differential privacy classification tree data publishing method (ACDP-Tree). In this paper, simulation experiments are carried out on real medical data sets. The experimental results show that the algorithm ensures the validity and availability of the data to a certain extent while ensuring that the patient's privacy is not leaked.

Securing federated learning with blockchain: a systematic literature review

Federated learning (FL) is a promising framework for distributed machine learning that trains models without sharing local data while protecting privacy. FL exploits the concept of collaborative learning and builds privacy-preserving models. Nevertheless, the integral features of FL are fraught with problems, such as the disclosure of private information, the unreliability of uploading model parameters to the server, the communication cost, etc. Blockchain, as a decentralized technology, is able to improve the performance of FL without requiring a centralized server and also solves the above problems. In this paper, a systematic literature review on the integration of Blockchain in federated learning was considered with the analysis of the existing FL problems that can be compensated. Through carefully screening, most relevant studies are included and research questions cover the potential security and privacy attacks in traditional federated learning that can be solved by blockchain as well as the characteristics of Blockchain-based FL. In addition, the latest Blockchain-based approaches to federated learning have been studied in-depth in terms of security and privacy, records and rewards, and verification and accountability. Furthermore, open issues related to the combination of Blockchain and FL are discussed. Finally, future research directions for the robust development of Blockchain-based FL systems are proposed.

A Deep Learning-Based Privacy-Preserving Model for Smart Healthcare in Internet of Medical Things Using Fog Computing.

With the emergence of COVID-19, smart healthcare, the Internet of Medical Things, and big data-driven medical applications have become even more important. The biomedical data produced is highly confidential and private. Unfortunately, conventional health systems cannot support such a colossal amount of biomedical data. Hence, data is typically stored and shared through the cloud. The shared data is then used for different purposes, such as research and discovery of unprecedented facts. Typically, biomedical data appear in textual form (e.g., test reports, prescriptions, and diagnosis). Unfortunately, such data is prone to several security threats and attacks, for example, privacy and confidentiality breach. Although significant progress has been made on securing biomedical data, most existing approaches yield long delays and cannot accommodate real-time responses. This paper proposes a novel fog-enabled privacy-preserving model called delta _r sanitizer, which uses deep learning to improve the healthcare system. The proposed model is based on a Convolutional Neural Network with Bidirectional-LSTM and effectively performs Medical Entity Recognition. The experimental results show that delta _r sanitizer outperforms the state-of-the-art models with 91.14% recall, 92.63% in precision, and 92% F1-score. The sanitization model shows 28.77% improved utility preservation as compared to the state-of-the-art.

Advanced Encryption Standard based Secure IoT Data Transfer Model for Cloud Analytics Applications

The data surge caused by the increase in the use of IoT devices in our day-to-day activities requires careful storage and maintenance systems to ensure the protection and integrity of private information. Users are provided with prospects to use rule-based programs and services that can be interconnected with their devices thereby enabling automation in the prodigious IoT ecosystem. The sensitive IoT data is generally stored and processed in cloud services. This data may be vulnerable to several threats. It is crucial to protect rule-based programs and sensitive IoT information against cyberattacks. The rule-based program and IoT data integrity and confidentiality can be maintained with the help of the framework proposed in this work. An end-to-end data encryption model based on advanced encryption standard and Intel SGX are used to design the framework of the data privacy preservation model. Real as well as simulated IoT device data are used for securely executing the rule-based programs in the SGX to evaluate the proposed framework.

Privacy-preserving Model Training for Disease Prediction Using Federated Learning with Differential Privacy.

Machine learning is playing an increasingly critical role in health science with its capability of inferring valuable information from high-dimensional data. More training data provides greater statistical power to generate better models that can help decision-making in healthcare. However, this often requires combining research and patient data across institutions and hospitals, which is not always possible due to privacy considerations. In this paper, we outline a simple federated learning algorithm implementing differential privacy to ensure privacy when training a machine learning model on data spread across different institutions. We tested our model by predicting breast cancer status from gene expression data. Our model achieves a similar level of accuracy and precision as a single-site non-private neural network model when we enforce privacy. This result suggests that our algorithm is an effective method of implementing differential privacy with federated learning, and clinical data scientists can use our general framework to produce differentially private models on federated datasets. Our framework is available at https://github.com/gersteinlab/idash20FL.

Bio inspired based Multi-objective Data-Privacy-Preservation-Technique (MO-DPPT) system model for cloud data

The significance of data security in the cloud system has boosted the increasing quantity of sensitive and personal data being harvested by data controllers. As the cloud has more outsourced, unsecured sensitive data for public access, the data security for the cloud sectors is very important. The majority of state-of-the-art strategies fail to manage optimal privacy in implementing data privacy preservation model. To ensure data privacy preservation, most of the traditional techniques can perform the transformation on the actual data. These traditional methods are utilized limitedly, as they are pretty memory intensive and complicated. Hence, to defeat this dispute, this paper tries to develop a novel approach named Multi-Objective Data-Privacy-Preservation-Technique (MO-DPPT) that can look after the data privacy issues. The two main phases of the data privacy preservation model consist of the data sanitization and restoration process. Here, the optimal key generation is based on the proposed sanitization process and it is done by a novel meta-heuristic algorithm namely the Muddy Soil Fish Optimization Algorithm (MSFOA). The optimal key generation is achieved by developing the multi-objective function that focuses on the parameters of hiding ratio (HR) rate, Preservation Ratio (PR) rate, False Rule (FR) generation rate, and Degree of Modification (DM). Moreover, the proposed MSFOA algorithm has confirmed improved execution through statistical analysis, analysis on KPA and CPA attacks, and computational time analysis over the conventional algorithms. Finally, the proposed MSFOA based MO-DPPT system model using is compared with the existing traditional algorithms namely PSO, GWO, JA, SSO, JA-SSO, and the performance analysis of the proposed MSFOA based MO-DPPT system model has proven the superior efficiency in improving cloud security.

Privacy Threats and Privacy Preservation Techniques for Farmer Data Collections Based on Data Shuffling

Aside from smart technologies, farm data collection is also important for smart farms including farm environment data collection and farmer survey data collection. With farm data collection, we observe that it is generally proposed to utilize in smart farm systems. However, it can also be released for use in the outside scope of the data collecting organization for an appropriate business reason such as improving the smart farm system, product quality, and customer service. Moreover, we can observe that the farmer survey data collection often includes sensitive data, the private data of farmers. Thus, it could lead to privacy violation issues when it is released. To address these issues in the farmer survey data collection, an anatomization model can protect the users' private data that is available in farmer survey data collection to be proposed. However, it still has disorganized issues and privacy violation issues in the sensitive table that must be addressed. To rid these vulnerabilities of anatomization models, a new privacy preservation model based on data shuffing is proposed in this work. Moreover, the proposed model is evaluated by conducting extensive experiments. The experimental results indicate that the proposed model is more efficient than the anatomization model for the farmer survey data collection. That is, the adversary can have the confidence for re-identifying every sensitive data that is available in farmer survey data collection that is after satisfied by the privacy preservation constraint of the proposed model to be at most 1/l. Furthermore, after the farmer survey data collection satisfies the privacy preservation constraint of the proposed model, it does not have disorganized issues and privacy violation issues from considering the sensitive values.

A federated graph neural network framework for privacy-preserving personalization

Graph neural network (GNN) is effective in modeling high-order interactions and has been widely used in various personalized applications such as recommendation. However, mainstream personalization methods rely on centralized GNN learning on global graphs, which have considerable privacy risks due to the privacy-sensitive nature of user data. Here, we present a federated GNN framework named FedPerGNN for both effective and privacy-preserving personalization. Through a privacy-preserving model update method, we can collaboratively train GNN models based on decentralized graphs inferred from local data. To further exploit graph information beyond local interactions, we introduce a privacy-preserving graph expansion protocol to incorporate high-order information under privacy protection. Experimental results on six datasets for personalization in different scenarios show that FedPerGNN achieves 4.0% ~ 9.6% lower errors than the state-of-the-art federated personalization methods under good privacy protection. FedPerGNN provides a promising direction to mining decentralized graph data in a privacy-preserving manner for responsible and intelligent personalization.

Energy-Aware Edge Association for Cluster-Based Personalized Federated Learning

Federated Learning (FL) over wireless network enables data-conscious services by leveraging the ubiquitous intelligence at network edge for privacy-preserving model training. As the proliferation of context-aware services, the diversified personal preferences cause disagreeing conditional distributions among user data, which leads to poor inference performance. In this sense, clustered federated learning is proposed to group user devices with similar preference and provide each cluster with a personalized model. This calls for innovative design in edge association that involves user clustering and also resource management optimization. We formulate an accuracy-cost trade-off optimization problem by jointly considering model accuracy, communication resource allocation and energy consumption. To comply with parameter encryption techniques in FL, we propose an iterative solution procedure which employs deep reinforcement learning based approach at cloud server for edge association. The reward function consists of minimized energy consumption at each base station and the averaged model accuracy of all users. Under our proposed solution, multiple edge base station are fully exploited to realize cost efficient personalized federated learning without any prior knowledge on model parameters. Simulation results show that our proposed strategy outperforms existing strategies in achieving accurate learning at low energy cost.