ETD

• data altruism
• Relational Database

The thesis addresses Data Altruism in the context of research infrastructures. The objective is to design and develop a prototype of a digital platform for the voluntary collection, management, and controlled reuse of data for objectives of general interest related to Data Altruism. The increasingly central role of data sharing in research and contemporary society highlights that the value of data depends on their availability and on mechanisms that ensure transparency, personal data protection, accountability, and trust among the actors involved.

The first part of the thesis analyses the relevant European regulatory framework, in particular the Data Governance Act and the GDPR. The DGA defines Data Altruism as the voluntary sharing of data, without reward, for objectives of general interest, including scientific research, public health, the environment, and public services. Since the processing of personal data remains subject to the GDPR, particular attention is devoted to consent, the possibility of withdrawal, transparency towards Data Subjects, the management of minors, roles and responsibilities, and constraints relating to data transfers.

Starting from this framework, the thesis proposes a platform model that could be inserted within the context of SoBigData Research Infrastructure, considered as a reference environment for responsible, ethical and privacy-aware data management. The model identifies four main roles: the Builder, who proposes and configures a Data Altruism Initiative; the Data Subject, who participates voluntarily, gives consent, and uploads their data; the Data User, who requests access to a specific Initiative's data through a coherent research project; and the Platform Administrator, who supervises the initiatives, evaluates access requests, and controls the correct execution of the workflows.

The technical part of the thesis translates the conceptual model into a relational database implemented in PostgreSQL that represents the main entities of the platform, including initiatives, consents, datasets, participating subjects, Data User projects, documents, access authorizations, and reports of incidental findings. A relevant aspect is the dynamic management of datasets: the Builder of each initiative can define a different dataset schema, and the system generates dedicated physical tables based on that schema. The platform also implements a pseudonymization mechanism that allows the separation of the Data Subject's identity from the uploaded data while maintaining the link necessary for operational management, including consent withdrawal.

The prototype also includes a middle layer in Python, responsible for input validation and executing the main database operations. The interface is developed with the Streamlit library and follows a role-based logic, in which each actor accesses only the functions consistent with their role. The platform therefore allows the creation and review of initiatives, the collection of consent, data upload, the submission of research projects, the upload of required documents, the approval of access, and the controlled download of datasets.

Finally, the thesis presents a complete simulation that shows the entire lifecycle of a Data Altruism Initiative, from its creation by the Builder to the participation of Data Subjects, the access request by the Data User and the supervision of the Platform Administrator.

Overall, the work demonstrates how the principles of Data Altruism can be translated into a working prototype that integrates legal requirements, data modeling, databases, application logic, and the user interface.