PROJECT TITLE :

Differentially Private Data Publishing and Analysis - 2017

ABSTRACT:

Differential privacy is an important and prevalent privacy model that has been widely explored in recent decades. This survey provides a comprehensive and structured overview of two analysis directions: differentially private data publishing and differentially non-public information analysis. We tend to compare the varied unleash mechanisms of differentially personal information publishing given a selection of input knowledge in terms of query sort, the maximum number of queries, potency, and accuracy. We have a tendency to establish two basic frameworks for differentially non-public knowledge analysis and list the standard algorithms used at intervals every framework. The results are compared and mentioned based mostly on output accuracy and potency. Any, we propose several doable directions for future research and doable applications.