Data sensitivity detection in chat interactions for privacy protection

In recent years, there has been exponential growth in using virtual spaces, including dialogue systems, that handle personal information. The concept of personal privacy in the literature is discussed and controversial, whereas, in the technological field, it directly influences the degree of reliability perceived in the information system (privacy ‘as trust’). This work aims to protect the right to privacy on personal data (GDPR, 2018) and avoid the loss of sensitive content by exploring sensitive information detection (SID) task. It is grounded on the following research questions: (RQ1) What does sensitive data mean? How to define a personal sensitive information domain? (RQ2) How to create a state-of-the-art model for SID?(RQ3) How to evaluate the model? RQ1 theoretically investigates the concepts of privacy and the ontological state-of-the-art representation of personal information. The Data Privacy Vocabulary (DPV) is the taxonomic resource taken as an authoritative reference for the definition of the knowledge domain. Concerning RQ2, we investigate two approaches to classify sensitive data: the first - bottom-up - explores automatic learning methods based on transformer networks, the second - top-down - proposes logical-symbolic methods with the construction of privaframe, a knowledge graph of compositional frames representing personal data categories. Both approaches are tested. For the evaluation - RQ3 – we create SPeDaC, a sentence-level labeled resource. This can be used as a benchmark or training in the SID task, filling the gap of a shared resource in this field. If the approach based on artificial neural networks confirms the validity of the direction adopted in the most recent studies on SID, the logical-symbolic approach emerges as the preferred way for the classification of fine-grained personal data categories, thanks to the semantic-grounded tailor modeling it allows. At the same time, the results highlight the strong potential of hybrid architectures in solving automatic tasks.

Governing algorithms in the big data era for balancing new digital rights - designing GDPR compliant and trustworthy XAI systems

This thesis investigates the legal, ethical, technical, and psychological issues of general data processing and artificial intelligence practices and the explainability of AI systems. It consists of two main parts. In the initial section, we provide a comprehensive overview of the big data processing ecosystem and the main challenges we face today. We then evaluate the GDPR’s data privacy framework in the European Union. The Trustworthy AI Framework proposed by the EU’s High-Level Expert Group on AI (AI HLEG) is examined in detail. The ethical principles for the foundation and realization of Trustworthy AI are analyzed along with the assessment list prepared by the AI HLEG. Then, we list the main big data challenges the European researchers and institutions identified and provide a literature review on the technical and organizational measures to address these challenges. A quantitative analysis is conducted on the identified big data challenges and the measures to address them, which leads to practical recommendations for better data processing and AI practices in the EU. In the subsequent part, we concentrate on the explainability of AI systems. We clarify the terminology and list the goals aimed at the explainability of AI systems. We identify the reasons for the explainability-accuracy trade-off and how we can address it. We conduct a comparative cognitive analysis between human reasoning and machine-generated explanations with the aim of understanding how explainable AI can contribute to human reasoning. We then focus on the technical and legal responses to remedy the explainability problem. In this part, GDPR’s right to explanation framework and safeguards are analyzed in-depth with their contribution to the realization of Trustworthy AI. Then, we analyze the explanation techniques applicable at different stages of machine learning and propose several recommendations in chronological order to develop GDPR-compliant and Trustworthy XAI systems.

EXAM-S: an Analysis tool for Multi-Domain Policy Sets

As distributed collaborative applications and architectures are adopting policy based management for tasks such as access control, network security and data privacy, the management and consolidation of a large number of policies is becoming a crucial component of such policy based systems. In large-scale distributed collaborative applications like web services, there is the need of analyzing policy interactions and integrating policies. In this thesis, we propose and implement EXAM-S, a comprehensive environment for policy analysis and management, which can be used to perform a variety of functions such as policy property analyses, policy similarity analysis, policy integration etc. As part of this environment, we have proposed and implemented new techniques for the analysis of policies that rely on a deep study of state of the art techniques. Moreover, we propose an approach for solving heterogeneity problems that usually arise when considering the analysis of policies belonging to different domains. Our work focuses on analysis of access control policies written in the dialect of XACML (Extensible Access Control Markup Language). We consider XACML policies because XACML is a rich language which can represent many policies of interest to real world applications and is gaining widespread adoption in the industry.

Integrated communication and learning towards autonomous mobile radio networks

The integration of distributed and ubiquitous intelligence has emerged over the last years as the mainspring of transformative advancements in mobile radio networks. As we approach the era of “mobile for intelligence”, next-generation wireless networks are poised to undergo significant and profound changes. Notably, the overarching challenge that lies ahead is the development and implementation of integrated communication and learning mechanisms that will enable the realization of autonomous mobile radio networks. The ultimate pursuit of eliminating human-in-the-loop constitutes an ambitious challenge, necessitating a meticulous delineation of the fundamental characteristics that artificial intelligence (AI) should possess to effectively achieve this objective. This challenge represents a paradigm shift in the design, deployment, and operation of wireless networks, where conventional, static configurations give way to dynamic, adaptive, and AI-native systems capable of self-optimization, self-sustainment, and learning. This thesis aims to provide a comprehensive exploration of the fundamental principles and practical approaches required to create autonomous mobile radio networks that seamlessly integrate communication and learning components. The first chapter of this thesis introduces the notion of Predictive Quality of Service (PQoS) and adaptive optimization and expands upon the challenge to achieve adaptable, reliable, and robust network performance in dynamic and ever-changing environments. The subsequent chapter delves into the revolutionary role of generative AI in shaping next-generation autonomous networks. This chapter emphasizes achieving trustworthy uncertainty-aware generation processes with the use of approximate Bayesian methods and aims to show how generative AI can improve generalization while reducing data communication costs. Finally, the thesis embarks on the topic of distributed learning over wireless networks. Distributed learning and its declinations, including multi-agent reinforcement learning systems and federated learning, have the potential to meet the scalability demands of modern data-driven applications, enabling efficient and collaborative model training across dynamic scenarios while ensuring data privacy and reducing communication overhead.

Application Platforms for the Internet of Things: Theory, Architecture, Protocols, Data Formats, and Privacy

The Internet of Things (IoT) is the next industrial revolution: we will interact naturally with real and virtual devices as a key part of our daily life. This technology shift is expected to be greater than the Web and Mobile combined. As extremely different technologies are needed to build connected devices, the Internet of Things field is a junction between electronics, telecommunications and software engineering. Internet of Things application development happens in silos, often using proprietary and closed communication protocols. There is the common belief that only if we can solve the interoperability problem we can have a real Internet of Things. After a deep analysis of the IoT protocols, we identified a set of primitives for IoT applications. We argue that each IoT protocol can be expressed in term of those primitives, thus solving the interoperability problem at the application protocol level. Moreover, the primitives are network and transport independent and make no assumption in that regard. This dissertation presents our implementation of an IoT platform: the Ponte project. Privacy issues follows the rise of the Internet of Things: it is clear that the IoT must ensure resilience to attacks, data authentication, access control and client privacy. We argue that it is not possible to solve the privacy issue without solving the interoperability problem: enforcing privacy rules implies the need to limit and filter the data delivery process. However, filtering data require knowledge of how the format and the semantics of the data: after an analysis of the possible data formats and representations for the IoT, we identify JSON-LD and the Semantic Web as the best solution for IoT applications. Then, this dissertation present our approach to increase the throughput of filtering semantic data by a factor of ten.

Data, Competition, and Consumer Privacy in Digital Markets

In digital markets personal information is pervasively collected by firms. In the first chapter I study data ownership and product customization when there is exclusive access to non rival but excludable data about consumer preferences. I show that an incumbent firm does not have an incentive to sell an exclusively held dataset with a rival firm, but instead it has an incentive to trade a customizing technology with the other firm. In the second chapter I investigate the effects of consumer information on the intensity of competition. In a two dimensional model of product differentiation, firms use information on preferences to practice price discrimination. I contrast a full privacy and a no privacy benchmark with a regime in which firms are able to target consumers only partially. When data is partially informative, firms are always better-off with price discrimination and an exclusive access to user data is not necessarily a competition policy concern. From a consumer protection perspective, the policy recommendation is that the regulator should promote either no privacy or full privacy. In the third chapter I introduce a data broker that observes either only one or both dimensions of consumer information and sells this data to competing firms for price discrimination purposes. When the seller exogenously holds a partially informative dataset, an exclusive allocation arises. Instead, when the dataset held is fully informative, the data broker trades information non exclusively but each competitor acquires consumer data on a different dimension. When data collection is made endogenous, non exclusivity is robust if collection costs are not too high. The competition policy suggestion is that exclusivity should not be banned per se, but it is data differentiation in equilibrium that rises market power in competitive markets. Upstream competition is sufficient to ensure that both firms get access to consumer information.

Big data in health IoE in emergency situations: between the right to privacy and digital health innovation

The chapters of the thesis focus on a limited variety of selected themes in EU privacy and data protection law. Chapter 1 sets out the general introduction on the research topic. Chapter 2 touches upon the methodology used in the research. Chapter 3 conceptualises the basic notions from a legal standpoint. Chapter 4 examines the current regulatory regime applicable to digital health technologies, healthcare emergencies, privacy, and data protection. Chapter 5 provides case studies on the application deployed in the Covid-19 scenario, from the perspective of privacy and data protection. Chapter 6 addresses the post-Covid European regulatory initiatives on the subject matter, and its potential effects on privacy and data protection. Chapter 7 is the outcome of a six-month internship with a company in Italy and focuses on the protection of fundamental rights through common standardisation and certification, demonstrating that such standards can serve as supporting tools to guarantee the right to privacy and data protection in digital health technologies. The thesis concludes with the observation that finding and transposing European privacy and data protection standards into scenarios, such as public healthcare emergencies where digital health technologies are deployed, requires rapid coordination between the European Data Protection Authorities and the Member States guarantee that individual privacy and data protection rights are ensured.

Big data analysis systems in IoE environments for managing privacy and data protection: pseudonymity, de-anonymization and the right to be forgotten

The thesis represents the conclusive outcome of the European Joint Doctorate programmein Law, Science & Technology funded by the European Commission with the instrument Marie Skłodowska-Curie Innovative Training Networks actions inside of the H2020, grantagreement n. 814177. The tension between data protection and privacy from one side, and the need of granting further uses of processed personal datails is investigated, drawing the lines of the technological development of the de-anonymization/re-identification risk with an explorative survey. After acknowledging its span, it is questioned whether a certain degree of anonymity can still be granted focusing on a double perspective: an objective and a subjective perspective. The objective perspective focuses on the data processing models per se, while the subjective perspective investigates whether the distribution of roles and responsibilities among stakeholders can ensure data anonymity.

Internet of healthcare (law): privacy and data protection aspects in an internet of everything

The purpose of this research study is to discuss privacy and data protection-related regulatory and compliance challenges posed by digital transformation in healthcare in the wake of the COVID-19 pandemic. The public health crisis accelerated the development of patient-centred remote/hybrid healthcare delivery models that make increased use of telehealth services and related digital solutions. The large-scale uptake of IoT-enabled medical devices and wellness applications, and the offering of healthcare services via healthcare platforms (online doctor marketplaces) have catalysed these developments. However, the use of new enabling technologies (IoT, AI) and the platformisation of healthcare pose complex challenges to the protection of patient’s privacy and personal data. This happens at a time when the EU is drawing up a new regulatory landscape for the use of data and digital technologies. Against this background, the study presents an interdisciplinary (normative and technology-oriented) critical assessment on how the new regulatory framework may affect privacy and data protection requirements regarding the deployment and use of Internet of Health Things (hardware) devices and interconnected software (AI systems). The study also assesses key privacy and data protection challenges that affect healthcare platforms (online doctor marketplaces) in their offering of video API-enabled teleconsultation services and their (anticipated) integration into the European Health Data Space. The overall conclusion of the study is that regulatory deficiencies may create integrity risks for the protection of privacy and personal data in telehealth due to uncertainties about the proper interplay, legal effects and effectiveness of (existing and proposed) EU legislation. The proliferation of normative measures may increase compliance costs, hinder innovation and ultimately, deprive European patients from state-of-the-art digital health technologies, which is paradoxically, the opposite of what the EU plans to achieve.

Il fascicolo sanitario elettronico tra Semantic Web e diritto alla privacy

Principale obiettivo della ricerca è quello di ricostruire lo stato dell’arte in materia di sanità elettronica e Fascicolo Sanitario Elettronico, con una precipua attenzione ai temi della protezione dei dati personali e dell’interoperabilità. A tal fine sono stati esaminati i documenti, vincolanti e non, dell’Unione europea nonché selezionati progetti europei e nazionali (come “Smart Open Services for European Patients” (EU); “Elektronische Gesundheitsakte” (Austria); “MedCom” (Danimarca); “Infrastruttura tecnologica del Fascicolo Sanitario Elettronico”, “OpenInFSE: Realizzazione di un’infrastruttura operativa a supporto dell’interoperabilità delle soluzioni territoriali di fascicolo sanitario elettronico nel contesto del sistema pubblico di connettività”, “Evoluzione e interoperabilità tecnologica del Fascicolo Sanitario Elettronico”, “IPSE - Sperimentazione di un sistema per l’interoperabilità europea e nazionale delle soluzioni di Fascicolo Sanitario Elettronico: componenti Patient Summary e ePrescription” (Italia)). Le analisi giuridiche e tecniche mostrano il bisogno urgente di definire modelli che incoraggino l’utilizzo di dati sanitari ed implementino strategie effettive per l’utilizzo con finalità secondarie di dati sanitari digitali , come Open Data e Linked Open Data. L’armonizzazione giuridica e tecnologica è vista come aspetto strategico per ridurre i conflitti in materia di protezione di dati personali esistenti nei Paesi membri nonché la mancanza di interoperabilità tra i sistemi informativi europei sui Fascicoli Sanitari Elettronici. A questo scopo sono state individuate tre linee guida: (1) armonizzazione normativa, (2) armonizzazione delle regole, (3) armonizzazione del design dei sistemi informativi. I principi della Privacy by Design (“prottivi” e “win-win”), così come gli standard del Semantic Web, sono considerate chiavi risolutive per il suddetto cambiamento.

Privacy Tradeoffs in Information Technology Law

The thesis aims to make the dynamics of the tradeoffs involving privacy more visible; both theoretically and in two of the central current policy debates in European data protection law, the right to be forgotten and online tracking. In doing so, it offers an explanation for data protection law from an economic perspective and provides a basis for the evaluation of further data protection measures.

Data protection by design in the e-health care sector: theoretical and applied perspectives

In the digital age, e-health technologies play a pivotal role in the processing of medical information. As personal health data represents sensitive information concerning a data subject, enhancing data protection and security of systems and practices has become a primary concern. In recent years, there has been an increasing interest in the concept of Privacy by Design, which aims at developing a product or a service in a way that it supports privacy principles and rules. In the EU, Article 25 of the General Data Protection Regulation provides a binding obligation of implementing Data Protection by Design technical and organisational measures. This thesis explores how an e-health system could be developed and how data processing activities could be carried out to apply data protection principles and requirements from the design stage. The research attempts to bridge the gap between the legal and technical disciplines on DPbD by providing a set of guidelines for the implementation of the principle. The work is based on literature review, legal and comparative analysis, and investigation of the existing technical solutions and engineering methodologies. The work can be differentiated by theoretical and applied perspectives. First, it critically conducts a legal analysis on the principle of PbD and it studies the DPbD legal obligation and the related provisions. Later, the research contextualises the rule in the health care field by investigating the applicable legal framework for personal health data processing. Moreover, the research focuses on the US legal system by conducting a comparative analysis. Adopting an applied perspective, the research investigates the existing technical methodologies and tools to design data protection and it proposes a set of comprehensive DPbD organisational and technical guidelines for a crucial case study, that is an Electronic Health Record system.

How do Firms ask for Consumers’ Data Permission? The Value of Companies Data Practices.

On May 25, 2018, the EU introduced the General Data Protection Regulation (GDPR) that offers EU citizens a shelter for their personal information by requesting companies to explain how people’s information is used clearly. To comply with the new law, European and non-European companies interacting with EU citizens undertook a massive data re-permission-request campaign. However, if on the one side the EU Regulator was particularly specific in defining the conditions to get customers’ data access, on the other side, it did not specify how the communication between firms and consumers should be designed. This has left firms free to develop their re-permission emails as they liked, plausibly coupling the informative nature of these privacy-related communications with other persuasive techniques to maximize data disclosure. Consequently, we took advantage of this colossal wave of simultaneous requests to provide insights into two issues. Firstly, we investigate how companies across industries and countries chose to frame their requests. Secondly, we investigate which are the factors that influenced the selection of alternative re-permission formats. In order to achieve these goals, we examine the content of a sample of 1506 re-permission emails sent by 1396 firms worldwide, and we identify the dominant “themes” characterizing these emails. We then relate these themes to both the expected benefits firms may derive from data usage and the possible risks they may experience from not being completely compliant to the spirit of the law. Our results show that: (1) most firms enriched their re-permission messages with persuasive arguments aiming at increasing consumers’ likelihood of relinquishing their data; (2) the use of persuasion is the outcome of a difficult tradeoff between costs and benefits; (3) most companies acted in their self-interest and “gamed the system”. Our results have important implications for policymakers, managers, and customers of the online sector.

Security and privacy of resource constrained devices

The thesis aims to present a comprehensive and holistic overview on cybersecurity and privacy & data protection aspects related to IoT resource-constrained devices. Chapter 1 introduces the current technical landscape by providing a working definition and architecture taxonomy of ‘Internet of Things’ and ‘resource-constrained devices’, coupled with a threat landscape where each specific attack is linked to a layer of the taxonomy. Chapter 2 lays down the theoretical foundations for an interdisciplinary approach and a unified, holistic vision of cybersecurity, safety and privacy justified by the ‘IoT revolution’ through the so-called infraethical perspective. Chapter 3 investigates whether and to what extent the fast-evolving European cybersecurity regulatory framework addresses the security challenges brought about by the IoT by allocating legal responsibilities to the right parties. Chapters 4 and 5 focus, on the other hand, on ‘privacy’ understood by proxy as to include EU data protection. In particular, Chapter 4 addresses three legal challenges brought about by the ubiquitous IoT data and metadata processing to EU privacy and data protection legal frameworks i.e., the ePrivacy Directive and the GDPR. Chapter 5 casts light on the risk management tool enshrined in EU data protection law, that is, Data Protection Impact Assessment (DPIA) and proposes an original DPIA methodology for connected devices, building on the CNIL (French data protection authority) model.

Patient data sharing for the improvement of healthcare and medical research: towards a moral duty of patients to share health-related data?

The project answers to the following central research question: ‘How would a moral duty of patients to transfer (health) data for the benefit of health care improvement, research, and public health in the eHealth sector sit within the existing confidentiality, privacy, and data protection legislations?’. The improvement of healthcare services, research, and public health relies on patient data, which is why one might raise the question concerning a potential moral responsibility of patients to transfer data concerning health. Such a responsibility logically would have subsequent consequences for care providers concerning the further transferring of health data with other healthcare providers or researchers and other organisations (who also possibly transfer the data further with others and other organisations). Otherwise, the purpose of the patients’ moral duty, i.e. to improve the care system and research, would be undermined. Albeit the arguments that may exist in favour of a moral responsibility of patients to share health-related data, there are also some moral hurdles that come with such a moral responsibility. Furthermore, the existing European and national confidentiality, privacy and data protection legislations appear to hamper such a possible moral duty, and they may need to be reconsidered to unlock the full use of data for healthcare and research.