A Systematic Review: Cluster based k-Anonymization Approaches for Big Data Privacy

Abstract

The growing adoption of Internet-based applications in diverse areas, such as marketing, analysis and research continuously expanded the volume of data, resulting in the accumulation of big data. Available Online information can endanger individuals by exposing their sensitive data to privacy risk.  Numerous privacy preservation techniques exist in literature including cryptography-based, differential privacy and k-anonymization based methods, are accessible to safeguard individual privacy. K-anonymity is an approach used to achieve data privacy by applying generalization and suppression to group of data by making them anonymous. This can be achieved using traditional way or via clustering algorithms. This paper presents an in-depth review of state-of-the-art cluster-based approaches for achieving k-anonymity of structural and stream based data. Moreover, detailed analysis of the effect of clustering approaches on measuring parameters such as data utility and information loss. It also discusses the challenges faced while using clustering techniques to grouping data for data k-anonymization