Date of Publication :14th November 2017

Abstract: A gigantic quantity of individual health information is accessible in modern decades and dispositioning of any part of this information establishes a huge risk in the field of healthcare. Enduring anonymization methods are only appropriate for single susceptible and low down dimensional data to remain with privacy particularly like generalization and bucketization. We propose an anonymization technique that is an amalgamation of the betterment of anatomization and improved slicing approach observing to the principle of k-anonymity and l-diversity for the reason of dealing with high dimensional data along with multiple susceptible data. The anatomization approach disrupts the correlation detected between the quasi-identifier attributes and susceptible attributes (SA) and turnouts’ two different tables with non-overlapping attributes. Hence, experimental outcomes specify that the suggested method can preserve the privacy of data with various sensitive attributes. The anatomization approach reduces the loss of information and slicing algorithm advices in the correlation preservation and usefulness which gives output in sinking the data dimensionality and information deficiency.

Reference :