Privacy-aware data-intensive applications

Abstract

The rise of Big Data is leading to an increasing demand for data-intensive applications (DIAs), which, in many cases, are expected to process massive amounts of sensitive data. In this context, ensuring data privacy becomes paramount. While the way we design and develop DIAs has radically changed over the last few years in order to deal with Big Data, there has been relatively little effort to make such design privacy-aware. As a result, enforcing privacy policies in large-scale data processing is currently an open research problem. This thesis proposal makes one step towards this investigation: after identifying the dataflow model as the reference computational model for large-scale DIAs, (1) we propose a novel language for specifying privacy policies on dataflow applications along with (2) a dataflow rewriting mechanism to enforce such policies during DIA execution. Although a systematic evaluation still needs to be carried out, preliminary results are promising. We plan to implement our approach within a model-driven solution to ultimately simplify the design and development of privacy-aware DIAs, i.e. DIAs that ensure privacy policies at runtime.