Session-dependent Usage Control for Big Data

Gabriele Baldi¹, Yair Diaz², Theo Dimitrakos², Fabio Martinelli¹, Christina Michailidou¹⁺,

Paolo Mori¹, Oleksii Osliak¹, and Andrea Saracino¹

¹Consiglio Nazionale delle Ricerche, Istituto Informatica e Telematica, Pisa, Italy

name.surname@iit.cnr.it

²German Research Center, Huawei Technologies Dusseldorf GmbH, Munich, Germany

name.surname@huawei.com

Abstract

Business strategies are increasingly driven by the integrated analysis of huge volumes of heterogeneous data, coming from different sources such for example social media or Internet of Things devices. The so called Big Data are considered as relevant assets by companies and organizations, since they can be analysed to create new valuable knowledge and insights that could help managers in their strategic decisions. The full potential of Big Data could be realized if the information was coming from several distinct sources, with different characteristics and target audience. Although, data producers are not always willing to share their data with other companies due to lack of trust and the absence of a data protection framework which can be adopted in a Big Data environment. In this work, we present BigUCON, a framework which exploits the Usage Control paradigm in order to provide an enhanced, expressive and flexible authorization support for data protection within the aforementioned environment. The framework is integrated in Apache Hadoop, a software library which provides the infrastructure for storing, mining and processing large data sets through a collection of open-source software.

Keywords: Distributed Big Data, Usage Control, Access Control, Hadoop

+: Corresponding author: Christina Michailidou

Consiglio Nazionale delle Ricerche, Istituto Informatica e Telematica, Pisa, Italy,
Tel: +393347649376

Journal of Internet Services and Information Security (JISIS), 10(3): 76-92, August 2020

DOI: 10.22667/JISIS.2020.08.31.076 [pdf]