National Technical University of Athens (NTUA)
Σημασιολογικές προσεγγίσεις στην προστασία της ιδιωτικότητας σε δεδομένα καταγραφών συμπεριφοράς στον ιστό
Abstract
dc:descriptionPrivacy preservation in data publishing has gained considerable attention during the last years due to the need of several organizations to share their data without revealing sensitive information about real persons or legal entities included in them. Representative examples of such datasets are the so-called customers' databases kept by enterprises, the data produced by clinical tests and experiments conducted in hospitals and related institutes, the query logs held by search engine providers like Google, the financial data from the public sector information systems, the social data produced by the participation of individuals in social networks like Facebook, the location data from telecom providers like Vodafone, and so forth. On the one hand, the records included in these datasets carry valuable information not only for their owners, but also for a plethora of enterprises, universities and institutions. The evaluation of the proposed information retrieval and data mining techniques through realistic experiments, the ability to perform large-scale market analysis, and the feasibility of conducting medical, social, and multidisciplinary studies based on these datasets are only a few examples that demonstrate the importance for the latter to be publicly available. On the other hand, taking into account that the actual data are produced by the activities of “real” people, which in the usual case are performed in private, they are likely to capture sensitive information about them (e.g. medical and financial information, political beliefs, sexual preferences, etc.) that can be disclosed even if they are published without the attributes that directly identify an entity; for instance, the person's name, the social ID number, or the IP of a web server. Even without these attributes, the disclosure of sensitive information about an individual, or an entity in general, is achieved when: (i) the “adversary” has a priori some form of background knowledge about the entity (e.g., knows a person's music preferences), and/or (ii) the published dataset is cross-checked with other publicly available data sources (e.g., demographic data, social profiles extracted from personal web pages, etc.).
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2015
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Liagouris, Ioannis
- Λιαγούρης, Ιωάννης
Subjects
dc:subject × 16- Προστασία της ιδιωτικότητας
- Δεδομένα αναζήτησης στον ιστό
- Ιατρικά δεδομένα
- Σημασιολογία
- Privacy
- Web search query logs
- Electronic healh records
- Semantics
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Natural Sciences
- Computer and Information Sciences
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/39783