Towards Privacy-Preserving Publication of Continuous and Dynamic Data Spatial Indexing and Bucketization Approaches
Year: 2013
Type: dissertation
Abstract: Privacy-Preserving Data Publishing (PPDP) has become a critical issue for companies and organizations that would release their data. Many organizations collect and distribute personal data for a variety of different purposes, including demographic and public health research. In these situations, the data distributor is often faced with a dilemma: how to publish this personal data for analysis purposes without endangering the privacy of the concerned individuals? Disseminating such information without the privacy scare is an important problem. On one hand, the data publishers need to protect the privacy of individuals and on the other hand, it is also extremely important to preserve the usefulness of the data for the researchers. In this dissertation, we mainly focus on crafting the notions of privacy in various settings. We show that spatial indexes are extremely efficient for data publication tasks due to their ability to scale up. An extensive empirical evaluation reveals that it is possible to disseminate high-quality data that follows meaningful notions of privacy. Furthermore, it is possible to do this efficiently for high dimensional very large data sets and this approach can be extended to stronger notions of privacy e.g., differential privacy. Also, sequential data is being increasingly employed in a wide variety of applications and the publication of sequential data is of utmost importance for the betterment of these applications. We provide a bucketization-based approach to achieve a stronger privacy guarantee along with higher utility of final release.
Source: HAL
Author: Adeel Anjum
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1579854257
