PRIVACY PRESERVING USING ANONYMIZATION AND PERTURBATION IN...
ABSTRACT
Data mining techniques are used for analysis purpose, but the data may contains sensitive information about
individuals, which individuals don’t want to be revealed, during data mining process. k anonymity is one of the technique, which
is used for preserving privacy in Data mining. In k anonymity, k records will appear similar in quasi identifier attribute. For
achieving this generalization or suppression can be used. In Generalization attribute values are replaced by less specific value and
in suppression attribute values are suppressed by meaningless characters like ‘*’ or ‘?’. In this paper we have proposed k
anonymity using suppression, crossover and perturbation in classification tree. Original dataset will be input to our algorithm
and anonymized data set is output, in anonymized data set number of tuple are same as original data set and we are comparing
accuracy of original dataset with anonymized dataset. Accuracy of anonymized dataset is better as compared to original data set.
