Automatic Selection of Sensitive Attributes in PPDP

Abstract

Privacy-preserving data publishing and data analysis conventional respectable work in past age as encouraging methodologies for sharing data as well as preserving privacy of individual. A remedy to this is data scrambling. Numerous algorithms are formulated to replace lawful information by fictitious but that is practical. Yet, nothing was proposed to automate the detection associated with information scrambling which is very much essential. In this paper, we propose a novel method for automatic selection of sensitive attributes for data scrambling. This is done by ranking the attributes based on attribute evaluation measures. And also, an innovative method for privacy-preserving data publishing with automatic selection of sensitive attributes which provide secure release of information for a data-mining task while preserving sensitive information.