Performance Evaluation of K-Anonymized Data

Abstract

Data mining provides tools to convert a large amount of knowledge data which is user relevant But this process could return individual s sensitive information compromising their privacy rights So based on different approaches many privacy protection mechanism incorporated data mining techniques were developed A widely used micro data protection concept is k-anonymity proposed to capture the protection of a micro data table regarding re-identification of respondents which the data refers to In this paper the effect of the anonymization due to k-anonymity on the data mining classifiers is investigated Na ve Bayes classifier is used for evaluating the anonymized and non-anonymized data