Privacy-preserving Technologies Research Articles

A Survey on Energy-Aware Security Mechanisms for the Internet of Things

The Internet of Things (IoT) employs sensors and the Internet for information exchange, enabling intelligent identification, monitoring, and management, which has deeply impacted various sectors such as power, medical care, and security, transforming social activities and lifestyles. Regrettably, IoT systems suffer from two main challenges, namely sustainability and security. Hence, pondering how to enhance sustainable and energy-efficient practices for IoT systems to mitigate risks becomes a worthwhile endeavor. To address this issue, we conduct a survey of energy-aware security mechanisms in the Internet of Things. Specifically, we examine the challenges that IoT is facing in terms of energy efficiency and security, and we inspect current energy-saving and privacy-preserving technologies for IoT systems. Moreover, we delineate a vision for the future of IoT, emphasizing energy-aware security mechanisms. Finally, we outline the challenges encountered in achieving energy-aware security mechanisms, as well as the direction of future research. Motivated by this study, we envision advancements in the IoT that not only harness the benefits of science and technology but also enhance the security and safety of our data.

Security and Privacy Issues in Deep Reinforcement Learning: Threats and Countermeasures

Deep Reinforcement Learning (DRL) is an essential subfield of Artificial Intelligence (AI), where agents interact with environments to learn policies for solving complex tasks. In recent years, DRL has achieved remarkable breakthroughs in various tasks, including video games, robotic control, quantitative trading, and autonomous driving. Despite its accomplishments, security and privacy-related issues still prevent us from deploying trustworthy DRL applications. For example, by manipulating the environment, an attacker can influence an agent’s actions, misleading it to behave abnormally. Additionally, an attacker can infer private training data and environmental information by maliciously interacting with DRL models, causing a privacy breach. In this survey, we systematically investigate the recent progress of security and privacy issues in the context of DRL. First, we present a holistic review of security-related attacks within DRL systems from the perspectives of single-agent and multi-agent systems and review privacy-related attacks. Second, we review and classify defense methods used to address security-related challenges, including robust learning, anomaly detection, and game theory approaches. Third, we review and classify privacy-preserving technologies, including encryption, differential privacy, and policy confusion. We conclude the survey by discussing open issues and possible directions for future research in this field.

Securing the virtual marketplace: Navigating the landscape of security and privacy challenges in E-Commerce

This paper provides a comprehensive examination of the evolving challenges and critical considerations surrounding security and privacy within the realm of e-commerce. As the digital marketplace continues to expand, the significance of safeguarding sensitive information and ensuring user privacy has become paramount. This paper explores various dimensions of security threats, including data breaches, phishing attacks, and vulnerabilities associated with payment gateways, shedding light on the potential repercussions for businesses and consumers alike. Additionally, it delves into emerging technologies and innovative solutions aimed at fortifying e-commerce platforms against evolving cyber threats. The paper not only identifies the existing vulnerabilities but also proposes proactive strategies and future directions for research and implementation. Topics such as biometric authentication, post-quantum cryptography, and privacy-preserving technologies are explored as potential avenues for enhancing the security posture of e-commerce systems. The research presented emphasizes the critical intersection of technology, regulation, and user awareness in fostering a secure and trustworthy online shopping environment. By offering insights into both the current state of e-commerce security and promising avenues for future exploration, this paper aims to contribute to the ongoing discourse on fortifying the digital marketplace against the challenges posed by an ever-evolving cyber landscape.

Eyes See Hazy while Algorithms Recognize Who You Are

Facial recognition technology has been developed and widely used for decades. However, it has also made privacy concerns and researchers’ expectations for facial recognition privacy-preserving technologies. To provide privacy, detailed or semantic contents in face images should be obfuscated. However, face recognition algorithms have to be tailor-designed according to current obfuscation methods, as a result the face recognition service provider has to update its commercial off-the-shelf (COTS) products for each obfuscation method. Meanwhile, current obfuscation methods have no clearly quantified explanation. This paper presents a universal face obfuscation method for a family of face recognition algorithms using global or local structure of eigenvector space. By specific mathematical explanations, we show that the upper bound of the distance between the original and obfuscated face images is smaller than the given recognition threshold. Experiments show that the recognition degradation is 0% for global structure based and 0.3%-5.3% for local structure based, respectively. Meanwhile, we show that even if an attacker knows the whole obfuscation method, he/she has to enumerate all the possible roots of a polynomial with an obfuscation coefficient, which is computationally infeasible to reconstruct original faces. So our method shows a good performance in both privacy and recognition accuracy without modifying recognition algorithms.

PICOTEES: a privacy-preserving online service of phenotype exploration for genetic-diagnostic variants from Chinese children cohorts

The growth in biomedical data resources has raised potential privacy concerns and risks of genetic information leakage. For instance, exome sequencing aids clinical decisions by comparing data through web services, but it requires significant trust between users and providers. To alleviate privacy concerns, the most commonly used strategy is to anonymize sensitive data. Unfortunately, studies have shown that anonymization is insufficient to protect against reidentification attacks. Recently, privacy-preserving technologies have been applied to preserve application utility while protecting the privacy of biomedical data. We present the PICOTEES framework, a privacy-preserving online service of phenotype exploration for genetic-diagnostic variants (https://birthdefectlab.cn:3000/). PICOTEES enables privacy-preserving queries of the phenotype spectrum for a single variant by utilizing trusted execution environment technology, which can protect the privacy of the user's query information, backend models, and data, as well as the final results. We demonstrate the utility and performance of PICOTEES by exploring a bioinformatics dataset. The dataset is from a cohort containing 20,909 genetic testing patients with 3,152,508 variants from the Children's Hospital of Fudan University in China, dominated by the Chinese Han population (>99.9%). Our query results yield a large number of unreported diagnostic variants and previously reported pathogenicity.

Evaluation Framework for the Use of Privacy Preserving Technologies for Production Data

To prevent unplanned machine downtime in production, machine conditions can be monitored and even predicted using condition and failure models based on current machine and process data. As most of these models are data-intensive, machine users often do not have enough data to develop these models themselves and want to collaborate with other companies. Since these models often require critical and classified machine and process data, which could be extracted from the models using attacks such as model inversion, sharing existing models between companies is not an option as it leaves one party vulnerable. Privacy preserving technologies such as homomorphic encryption, differential privacy, federated learning and secure multi-party computation can help overcome this problem. With the help of these approaches, there is no need to transmit sensitive data unencrypted to third parties in order to cooperate and take advantage of high-performance models. The aim of this paper is to first summarize the current state of research on privacy-preserving technologies in production, and then to provide a simple to use evaluation method and criteria. The focus is on enabling production workers to make informed decisions and exploit the full potential of existing data without the need for prior knowledge of privacy-preserving technologies. Finally, the evaluation method is validated using two example use cases in a production environment and the results are discussed.

Federated Learning for Clients’ Data Privacy Assurance in Food Service Industry

The food service industry must ensure that service facilities are free of foodborne pathogens hosted by organic residues and biofilms. Foodborne diseases put customers at risk and compromise the reputations of service providers. Fluorescence imaging, empowered by state-of-the-art artificial intelligence (AI) algorithms, can detect invisible residues. However, using AI requires large datasets that are most effective when collected from actual users, raising concerns about data privacy and possible leakage of sensitive information. In this study, we employed a decentralized privacy-preserving technology to address client data privacy issues. When federated learning (FL) is used, there is no need for data sharing across clients or data centralization on a server. We used FL and a new fluorescence imaging technology and applied two deep learning models, MobileNetv3 and DeepLabv3+, to identify and segment invisible residues on food preparation equipment and surfaces. We used FedML as our FL framework and Fedavg as the aggregation algorithm. The model achieved training and testing accuracies of 95.83% and 94.94% for classification between clean and contamination frames, respectively, and resulted in intersection over union (IoU) scores of 91.23% and 89.45% for training and testing, respectively, of segmentation of the contaminated areas. The results demonstrated that using federated learning combined with fluorescence imaging and deep learning algorithms can improve the performance of cleanliness auditing systems while assuring client data privacy.

Comparing Decentralized Learning Methods for Health Data Models to Nondecentralized Alternatives: Protocol for a Systematic Review.

Considering the soaring health-related costs directed toward a growing, aging, and comorbid population, the health sector needs effective data-driven interventions while managing rising care costs. While health interventions using data mining have become more robust and adopted, they often demand high-quality big data. However, growing privacy concerns have hindered large-scale data sharing. In parallel, recently introduced legal instruments require complex implementations, especially when it comes to biomedical data. New privacy-preserving technologies, such as decentralized learning, make it possible to create health models without mobilizing data sets by using distributed computation principles. Several multinational partnerships, including a recent agreement between the United States and the European Union, are adopting these techniques for next-generation data science. While these approaches are promising, there is no clear and robust evidence synthesis of health care applications. The main aim is to compare the performance among health data models (eg, automated diagnosis and mortality prediction) developed using decentralized learning approaches (eg, federated and blockchain) to those using centralized or local methods. Secondary aims are comparing the privacy compromise and resource use among model architectures. We will conduct a systematic review using the first-ever registered research protocol for this topic following a robust search methodology, including several biomedical and computational databases. This work will compare health data models differing in development architecture, grouping them according to their clinical applications. For reporting purposes, a PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 flow diagram will be presented. CHARMS (Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies)-based forms will be used for data extraction and to assess the risk of bias, alongside PROBAST (Prediction Model Risk of Bias Assessment Tool). All effect measures in the original studies will be reported. The queries and data extractions are expected to start on February 28, 2023, and end by July 31, 2023. The research protocol was registered with PROSPERO, under the number 393126, on February 3, 2023. With this protocol, we detail how we will conduct the systematic review. With that study, we aim to summarize the progress and findings from state-of-the-art decentralized learning models in health care in comparison to their local and centralized counterparts. Results are expected to clarify the consensuses and heterogeneities reported and help guide the research and development of new robust and sustainable applications to address the health data privacy problem, with applicability in real-world settings. We expect to clearly present the status quo of these privacy-preserving technologies in health care. With this robust synthesis of the currently available scientific evidence, the review will inform health technology assessment and evidence-based decisions, from health professionals, data scientists, and policy makers alike. Importantly, it should also guide the development and application of new tools in service of patients' privacy and future research. PROSPERO 393126; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=393126. PRR1-10.2196/45823.

Heterogeneous graph neural network with semantic-aware differential privacy guarantees

Most social networks can be modeled as heterogeneous graphs. Recently, advanced graph learning methods exploit the rich node properties and topological relationships for downstream tasks. That means that more private information is embedded in the representation. However, the existing privacy-preserving methods only focus on protecting the single type of node attributes or relationships, which neglect the significance of high-order semantic information. To address this issue, we propose a novel Heterogeneous graph neural network with Semantic-aware Differential privacy Guarantees named HeteSDG, which provides a double privacy guarantee and performance trade-off in terms of both graph features and topology. In particular, we first reveal the privacy leakage in heterogeneous graphs and define a membership inference attack with a semantic enhancement (MIS) that will improve the means of member inference attacks by obtaining side background knowledge through semantics. Then we design a two-stage mechanism, which includes the feature attention personalized mechanism and the topology gradient perturbation mechanism, where the privacy-preserving technologies are based on differential privacy. These mechanisms will defend against MIS and provide stronger interpretation, but simultaneously bring in noise for representation learning. To better balance the noise perturbation and learning performance, we utilize a bi-level optimization pattern to allocate a suitable privacy budget for the above two modules. Our experiments on four public benchmarks conduct performance experiments, ablation studies, inference attack verification, etc. The results show the privacy protection capability and generalization of HeteSDG.

Recent Trends on Privacy-Preserving Technologies under Standardization at the IETF

End-users are concerned about protecting the privacy of their sensitive personal data that are generated while working on information systems. This extends to both the data they actively provide including personal identification in exchange for products and services as well as its related metadata such as unnecessary access to their location. This is when certain privacy-preserving technologies come into a place where Internet Engineering Task Force (IETF) plays a major role in incorporating such technologies at the fundamental level. Thus, this paper offers an overview of the privacy-preserving mechanisms for layer 3 (i.e. IP) and above that are currently under standardization at the IETF. This includes encrypted DNS at layer 5 classified as DNS-over-TLS (DoT), DNS-over-HTTPS (DoH), and DNS-over-QUIC (DoQ) where the underlying technologies like QUIC belong to layer 4. Followed by that, we discuss Privacy Pass Protocol and its application in generating Private Access Tokens and Passkeys to replace passwords for authentication at the application layer (i.e. end-user devices). Lastly, to protect user privacy at the IP level, Private Relays and MASQUE are discussed. This aims to make designers, implementers, and users of the Internet aware of privacy-related design choices.

Exploring the Relationship Between Privacy and Utility in Mobile Health: Algorithm Development and Validation via Simulations of Federated Learning, Differential Privacy, and External Attacks

Although evidence supporting the feasibility of large-scale mobile health (mHealth) systems continues to grow, privacy protection remains an important implementation challenge. The potential scale of publicly available mHealth applications and the sensitive nature of the data involved will inevitably attract unwanted attention from adversarial actors seeking to compromise user privacy. Although privacy-preserving technologies such as federated learning (FL) and differential privacy (DP) offer strong theoretical guarantees, it is not clear how such technologies actually perform under real-world conditions. Using data from the University of Michigan Intern Health Study (IHS), we assessed the privacy protection capabilities of FL and DP against the trade-offs in the associated model's accuracy and training time. Using a simulated external attack on a target mHealth system, we aimed to measure the effectiveness of such an attack under various levels of privacy protection on the target system and measure the costs to the target system's performance associated with the chosen levels of privacy protection. A neural network classifier that attempts to predict IHS participant daily mood ecological momentary assessment score from sensor data served as our target system. An external attacker attempted to identify participants whose average mood ecological momentary assessment score is lower than the global average. The attack followed techniques in the literature, given the relevant assumptions about the abilities of the attacker. For measuring attack effectiveness, we collected attack success metrics (area under the curve [AUC], positive predictive value, and sensitivity), and for measuring privacy costs, we calculated the target model training time and measured the model utility metrics. Both sets of metrics are reported under varying degrees of privacy protection on the target. We found that FL alone does not provide adequate protection against the privacy attack proposed above, where the attacker's AUC in determining which participants exhibit lower than average mood is over 0.90 in the worst-case scenario. However, under the highest level of DP tested in this study, the attacker's AUC fell to approximately 0.59 with only a 10% point decrease in the target's R2 and a 43% increase in model training time. Attack positive predictive value and sensitivity followed similar trends. Finally, we showed that participants in the IHS most likely to require strong privacy protection are also most at risk from this particular privacy attack and subsequently stand to benefit the most from these privacy-preserving technologies. Our results demonstrated both the necessity of proactive privacy protection research and the feasibility of the current FL and DP methods implemented in a real mHealth scenario. Our simulation methods characterized the privacy-utility trade-off in our mHealth setup using highly interpretable metrics, providing a framework for future research into privacy-preserving technologies in data-driven health and medical applications.

The Paradoxical Impact of Information Privacy on Privacy Preserving Technology: The Case of Self-Sovereign Identities

Advance of digital technologies brings great benefits but takes users at risk of the dark sides of the internet. Preventive mechanisms and privacy-preserving solutions could overcome this challenge. As such, self-sovereign identities (SSIs) provide users with increased control over personal information. However, users neglect their privacy in favor of the most convenient solution. In this paper, we empirically examine how information privacy influences adoption of SSIs. Our results contradict the existing theory that privacy is critical to the success of identity management (IdM) systems. Analogous to the privacy paradox, the study does not lend empirical support that perceived privacy has an impact on the adoption of an SSI. On the contrary, these findings contradict the prevailing view of privacy as a key factor for IdM systems and contribute to knowledge on privacy and adoption behavior.

Security and Privacy Aware Programming Model for IoT Applications in Cloud Environment

The introduction of Internet of Things (IoT) applications into daily life has raised serious privacy concerns among consumers, network service providers, device manufacturers, and other parties involved. This paper gives a high-level overview of the three phases of data collecting, transmission, and storage in IoT systems as well as current privacy-preserving technologies. The following elements were investigated during these three phases:(1) Physical and data connection layer security mechanisms(2) Network remedies(3) Techniques for distributing and storing data. Real-world systems frequently have multiple phases and incorporate a variety of methods to guarantee privacy. Therefore, for IoT research, design, development, and operation, having a thorough understanding of all phases and their technologies can be beneficial. In this Study introduced two independent methodologies namely generic differential privacy (GenDP) and Cluster-Based Differential privacy ( Cluster-based DP) algorithms for handling metadata as intents and intent scope to maintain privacy and security of IoT data in cloud environments. With its help, we can virtual and connect enormous numbers of devices, get a clearer understanding of the IoT architecture, and store data eternally. However, due of the dynamic nature of the environment, the diversity of devices, the ad hoc requirements of multiple stakeholders, and hardware or network failures, it is a very challenging task to create security-, privacy-, safety-, and quality-aware Internet of Things apps. It is becoming more and more important to improve data privacy and security through appropriate data acquisition. The proposed approach resulted in reduced loss performance as compared to Support Vector Machine (SVM) , Random Forest (RF) .

IOT DEVICE SECURITY RISKS: A COMPREHENSIVE OVERVIEW AND MITIGATION STRATEGIES

The research paper delves into the intricate realm of IoT device security, unravelling the multifaceted risks and presenting a nuanced exploration of mitigation strategies. A comprehensive literature review unveils common security threads and the current state of IoT security measures. The subsequent analysis identifies security risks, including unauthorized access, encryption lapses, authentication weaknesses, physical vulnerabilities, and privacy concerns. Mitigation strategies encompass technical measures, policy frameworks, and user education, forming a holistic approach. The paper concludes by outlining recommendations for future research, emphasizing interdisciplinary collaboration, dynamic threat modelling, privacy-preserving technologies, standardization, certification, and blockchain integration. Envisioning a secure and connected future, the research underscores the pivotal role of manufacturers, policymakers, and users in shaping a resilient IoT landscape.

Digital Health Solutions for COVID-19.

In May 2020, in response to the rapidly expanding pandemic caused by the SARS-CoV-2 virus, the National Cancer Institute (NCI) and the National Institute of Biomedical Imaging and Bioengineering (NIBIB) together launched a program called “Digital Health Solutions for COVID-19.” The goal of this initiative was to support the development of digital health tools that could leverage multiple data sources, privacy-preserving technologies, and computational tools to assist in managing population health during the COVID-19 pandemic. Seven projects received National Institutes of Health (NIH) support to develop a variety of smartphone apps, wearable devices, and software tools capable of identifying infected individuals, tracking verified COVID-19 test results, and monitoring the health status of infected and potentially infected individuals. The NIH also supported the establishment of a data hub to facilitate the collection and curation of de-identified data generated by these projects, which will be made publicly available in late 2022. This will provide an opportunity for the broader research community to further analyze the data, refine predictive models, and continue to generate new learnings from the data sets. A major focus of the NCI/NIBIB initiative was to define the context of use for various digital health technologies to determine what could be learned from the data and how this could be leveraged to inform decision making during future pandemics. Among these technologies was an artificial intelligence, web-based decision aid, developed by Vibrent Health (Fairfax, Virginia), which can be used to evaluate COVID-19 symptoms of patients at home. Several articles in this issue describe how this system, and potentially others like it, can be extremely useful for both triage decisions and improving accuracy of in-home testing. While these testing improvements have been invaluable during the pandemic, such testing enhancements will become even more critical as COVID-19 shifts to an endemic. It will no longer be acceptable to triage patients based on general symptoms, such as fever and cough, as was done early in the pandemic. As patients return to normal activities and continue striving to make informed health care decisions, they will rightly expect more sophisticated methods to differentiate COVID-19 from other diseases with overlapping symptomatology. Accomplishing this goal will be tremendously complex and will require painstaking deconvolution of symptoms, but this can be achieved with the help of machine learning, as is shown in the Vibrent studies. Despite the fear and uncertainty throughout the COVID-19 pandemic, the public has demonstrated a remarkable willingness to participate in NIH-supported efforts aimed at improving COVID-19 outcomes. Projects supported under the Digital Health Solutions for COVID-19 initiative began in September 2020, just before the beta wave of the pandemic began, and study enrollment was generally high throughout the winter of 2020-2021. Study enrollment decreased somewhat in early 2021 with the rollout of COVID-19 vaccines and falling case numbers, and then steadily increased again during the spread of the delta variant in mid-2021. To successfully navigate these changing dynamics, it was especially important for the NIH-funded teams to establish clear and consistent communication with study participants. The team at Vibrent Health was able to successfully achieve this through targeted outreach campaigns and thoughtful community engagement. Particularly important were their studies aimed at engaging traditionally underserved and underrepresented populations, which showed that digital health tools play an important role in reaching populations that are not well served by current health systems. The Digital Health Solutions for COVID-19 initiative was just one component of a major coordinated effort by the NIH to combat COVID-19. Other programs such as the Rapid Acceleration of Diagnostics (RADx) Tech initiative also played a key role in speeding the development, validation, and commercializaiton of rapid, low-cost, point-of-care, and home-based tests. Importantly, large investments in next-generation testing approaches can only realize their full potential if these new technologies can be easily accessed and operated by the intended end users. The solutions developed by Vibrent Health and the other project teams have clearly demonstrated that digital health tools will play a crucial role in enhancing the use of various physical technologies going forward. In the future, public health officials, clinicians, employers, and the general community will rely heavily on these digital tools as they work to provide individuals with the necessary data to make informed choices about their health. The recent investment by the US government in digital health solutions and other COVID-19 technologies has not only provided clear and immediate benefits but also helped secure a solid foundation and infrastructure for the early detection, prevention, and monitoring of future pandemics. Hopefully, the recent COVID-19 experience will not soon be repeated, but the lessons learned will be invaluable when the next pandemic does occur. —Andrew Kurtz, PhD Program Director Center for Strategic Scientific Initiatives National Cancer Institute Bethesda, Maryland

Privacy-Preserving Contrastive Explanations with Local Foil Trees

We present the first algorithm that combines privacy-preserving technologies and state-of-the-art explainable AI to enable privacy-friendly explanations of black-box AI models. We provide a secure algorithm for contrastive explanations of black-box machine learning models that securely trains and uses local foil trees. Our work shows that the quality of these explanations can be upheld whilst ensuring the privacy of both the training data and the model itself.

The business value of privacy-preserving technologies: the case of multiparty computation in the telecom industry

PurposeThis paper aims to propose a framework for how privacy-preserving technologies (PETs) create business value for organizations. The framework was developed by examining the literature on privacy and information technology’s impact (symbolic and function). The authors evaluate the framework’s applicability using multiparty computation (MPC) as an instance of PETs, with expert interviews in the telecommunication industry.Design/methodology/approachIn an illustrative case of four telecommunication companies, the authors conducted semi-structured interviews with experts and used MPC as an instance of PET.FindingsThe evaluation of the framework indicates that PETs create business value for organizations: enhancing customer interactions, sales, personalized services, predicting market trends and collaboration among organizations. The findings show that business value of PETs is mainly driven by consumers and organizations willing to share data and collaborate.Research limitations/implicationsThis study was limited to the telecom sector and focused on MPC as an instance of PET. Further studies should be conducted to explore the benefits of other PETs and MPC. Future research could find out if this framework is also helpful for implementing other PETs or even other types of technology. The authors’ framework provides factors that future studies can use to quantify the impact of PETs. The authors hope that this framework provides an overarching reference for organizations considering the adoption of PETs.Practical implicationsThe authors’ findings inform managers in exploring the business value of PETs for organizations. This study also provides insights into which costs and risks to consider when implementing PETs.Originality/valueThis study is one of the few to propose a framework on how PETs create business value for organizations. Future research can use factors in the framework (e.g. customer interactions, sales, personalized services and market trend prediction) to conduct a quantitative study on PETs’ business value. Managers adopting PETs can use the framework to identify areas where PETs impact their organization.

Record linkage based patient intersection cardinality for rare disease studies using Mainzelliste and secure multi-party computation

BackgroundThe low number of patients suffering from any given rare diseases poses a difficult problem for medical research: With the exception of some specialized biobanks and disease registries, potential study participants’ information are disjoint and distributed over many medical institutions. Whenever some of those facilities are in close proximity, a significant overlap of patients can reasonably be expected, further complicating statistical study feasibility assessments and data gathering. Due to the sensitive nature of medical records and identifying data, data transfer and joint computations are often forbidden by law or associated with prohibitive amounts of effort. To alleviate this problem and to support rare disease research, we developed the Mainzelliste Secure EpiLinker (MainSEL) record linkage framework, a secure Multi-Party Computation based application using trusted-third-party-less cryptographic protocols to perform privacy-preserving record linkage with high security guarantees. In this work, we extend MainSEL to allow the record linkage based calculation of the number of common patients between institutions. This allows privacy-preserving statistical feasibility estimations for further analyses and data consolidation. Additionally, we created easy to deploy software packages using microservice containerization and continuous deployment/continuous integration. We performed tests with medical researchers using MainSEL in real-world medical IT environments, using synthetic patient data.ResultsWe show that MainSEL achieves practical runtimes, performing 10 000 comparisons in approximately 5 minutes. Our approach proved to be feasible in a wide range of network settings and use cases. The “lessons learned” from the real-world testing show the need to explicitly support and document the usage and deployment for both analysis pipeline integration and researcher driven ad-hoc analysis use cases, thus clarifying the wide applicability of our software. MainSEL is freely available under: https://github.com/medicalinformatics/MainSELConclusionsMainSEL performs well in real-world settings and is a useful tool not only for rare disease research, but medical research in general. It achieves practical runtimes, improved security guarantees compared to existing solutions, and is simple to deploy in strict clinical IT environments. Based on the “lessons learned” from the real-word testing, we hope to enable a wide range of medical researchers to meet their needs and requirements using modern privacy-preserving technologies.

ADFL: A Poisoning Attack Defense Framework for Horizontal Federated Learning

Recently, federated learning has received widespread attention, which will promote the implementation of artificial intelligence technology in various fields. Privacy-preserving technologies are applied to users’ local models to protect users’ privacy. Such operations make the server not see the true model parameters of each user, which opens wider door for a malicious user to upload malicious parameters and make the training result converge to an ineffective model. To solve this problem, in this article, we propose a poisoning attack defense framework for horizontal federated learning systems called ADFL. Specifically, we design a proof generation method for users to generate proofs to verify whether it is malicious or not. An aggregation rule is also proposed to make sure the global model has a high accuracy. Several verification experiments were conducted and the results show that our method can detect malicious user effectively and ensure the global model has a high accuracy.

Data Sharing Made Easy by Technology Trends: New Data Sharing and Privacy Preserving Technologies that Bring in a New Era of Data Monetization

Data Sharing Made Easy by Technology Trends: New Data Sharing and Privacy Preserving Technologies that Bring in a New Era of Data Monetization