Προστασία ιδιωτικότητας σε δεδομένα κίνησης
Privacy protection in mobility data
Γκάτσου, Ευγενία Η.
SubjectΙδιωτικότητα ; Γεωγραφικά Συστήματα Πληροφοριών -- Στατιστικές μέθοδοι ; Στατιστική -- Οικονομετρικές μέθοδοι
This thesis is focused on the subject of privacy protection in mobility data, examining methods that aim to protect sensitive user information by offering ways of anonymity. As an example of the necessity of these methods, we can mention GPS users, who in order to obtain the information they need, disclose their personal data. The present work focuses on two main themes, the literature review and Privacy by Design. In the first part, the introduction of the thesis is presented. We analyze some basic concepts that are helpful for the progress of the work. There is also a presentation of the contents of the subsequent chapters. The second part focuses on techniques that provide data protection. The model of k-anonymity is the most popular technique; based on which many other approaches were “born”. This model, in order to disclose data, requires the separation into equivalent classes sized at least k, where each class records are indistinguishable to sensitive attributes. In an attempt to overcome the limitations of k-anonymity and to provide a stronger sense of privacy, the models of l-diversity and t-closeness were introduced. In this part, basic privacy protection techniques in spatial and spatiotemporal data are also presented. The third part is referred to the general principle of Privacy by Design. We analyze the three assumptions and the seven fundamental principles needed to be followed by all techniques, in order to achieve an optimal trade-off between privacy and data utility. In addition, we check whether the most significant privacy protection techniques comply with the rules. In the fourth and final part, we present the conclusions of the entire thesis and our suggestions for future work.