Unit:
Κατεύθυνση / ειδίκευση Υπολογιστικά Συστήματα: Λογισμικό και Υλικό (ΣΥΣ)Library of the School of Science
Author:
Χατζόπουλος Σεραφείμ
Supervisors info:
Μέμα Ρουσσοπούλου
Original Title:
Survey of Privacy-Preserving Data Publishing Methods and Speedy: a multi-threaded algorithm preserving k-anonymity
Translated title:
Βιβλιογραφική επισκόπηση μεθόδων προστασίας της ιδιωτικότητας δεδομένων προς δημοσίευση και Speedy: ένας πολυνηματικός αλγόριθμος που διαφυλάσσει την k-ανωνυμία
Summary:
Nowadays, many organizations, enterprises or public services collect and manage
a vast amount of personal information. Typical examples of such datasets
include clinical tests conducted in hospitals, query logs held by search
engines, social data produced by social networks, financial data from public
sector information systems etc. These datasets often need to be published for
research or statistical studies without revealing sensitive information of the
individuals they describe. The anonymization process is more complicated than
hiding attributes that can directly identify an individual (name, SSN etc.)
from the published dataset. Even without these attributes an adversary can
cause privacy leakage by cross-linking with other publicly available datasets
or having some sort of background knowledge. Therefore, privacy preservation in
data publishing has gained considerable attention during recent years with
several privacy models proposed in the literature. In this thesis, we discuss
the most common attacks that can be made on published datasets and we present
state-of-the-art privacy guarantees and anonymization algorithms to counter
these attacks. Furthermore, we propose a novel multi-threaded anonymization
algorithm which exploits the capabilities of modern CPUs to speed up the
anonymization process achieving k-anonymity in the anonymized dataset.
Keywords:
privacy preservation, data anonymity, database systems, k-anonymity, multi-threaded algorithm
Number of index pages:
8-12
document.pdf (2 MB)
Open in new window
attachments.zip
57 KB
File access is restricted.
