Data quality and privacy action

Abstract

In an era where the volume of data sets for a particular topic keeps increasing, it is necessary to have quality checks on the data to keep it relevant to modern learning algorithms. Quality parameters provide an objective way to look at a data set and compare it with other data sets in similar domain. With many of the data sets becoming public, it is necessary to check whether they could compromise the privacy of individuals. This work explored topic agnostic and topic dependent quality parameters for the medical domain. Following this, we built a Data-Sharing Platform with insights to help facilitate data exchange (which is NDHM Compliant) between parties with appropriate consents.