Detection and Privacy Preservation of Sensitive Attributes Using Hybrid Approach for Privacy Preserving Record Linkage

Abstract

Privacy Preserving Record Linkage (PPRL) is a major area of database research which entangles in colluding huge multiple heterogeneous data sets with disjunctive or additional information about an entity while veiling its private information. This paper gives an enhanced algorithm for merging two datasets using Sorted Neighborhood Deterministic approach and an improved Preservation algorithm which makes use of automatic selection of sensitive attributes and pattern mining over dynamic queries. We guarantee strong privacy, less computational complexity and scalability and address the legitimate concerns over data security and privacy with our approach.