Year: 2013
Type: dissertation
Abstract: Data-as-a-Service (DaaS) is a paradigm that provides data on demand to consumers \nacross different cloud platforms over the Internet. Yet, a single DaaS provider may not be \nable to fulfill a data request. Consequently, the concept of DaaS mashup was introduced to \nenable DaaS providers to dynamically integrate their data on demand depending on consumers’ \nrequests. Utilizing DaaS mashup, however, involves some challenges. Mashing \nup data from multiple sources to answer a consumer’s request might reveal sensitive information \nand thereby compromise the privacy of individuals. Moreover, data integration \nof arbitrary DaaS providers might not always be sufficient to answer incoming requests. \nIn this thesis, we provide a cloud-based framework for privacy-preserving DaaS mashup \nthat enables secure collaboration between DaaS providers for the purpose of generating an \nanonymous dataset to support data mining. We propose a greedy algorithm to determine \na suitable group of DaaS providers whose data can satisfy a given request. Furthermore, \nour framework securely integrates the data from multiple DaaS providers while preserving \nthe privacy of the resulting mashup data. Experiments on real-life data demonstrate that \nour DaaS mashup framework is scalable to large set of databases and it can efficiently and effectively satisfy the data privacy and data mining requirements specified by the DaaS \nproviders and the data consumers.
Author: Mahtab Arafati
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1634670038