An Efficient Way of Anonymization Without Subjecting to Attacks Using Secure Matrix Method
T. Satyanarayana Murthy, N. P. Gopalan, G. Preethi · 2018
In current times huge data evolving from multiple sources like hospitals, reservation agencies, online transactions, etc. in massive volumes and obtaining in various forms. These data have privacy concerns due to leakage of data. The outgrowths raised in this situation may drive towards anonymization of sensitive identity information. Let a dataset released for the research purpose by removing the identifying attributes and sensitive attributes, but an adversary find to disclose the identity of the individuals by using the quasi-identifiers and non-sensitive data. Anonymization methods are classified into k-Anonymity, 1-diversity, and t-closeness fail in the better way of hiding the data. These techniques lead to a homogeneous attack, background knowledge attack, and similarity attack. In this article, novel method has been proposed based on secure matrix methods for an effective way of hiding the critical data. This technique accepts the non identified data as an input and produces anonymized data as an output without subjecting to attacks. It experimentally produces better results in anonymizing the data with less execution time.