Privacy Preserving Decision Tree Mining based on Mixed Attributes

Abstract

Privacy preservation is an important concept in the field of data mining approach to introduce the protection of dataset. Our work is to secure the datasets that are utilized for the data mining and mainly focuses on conversion of original datasets into unrealized datasets. During the conversion of unrealized datasets if any leakage occurs it should be recovered by cryptographic technique i.e., Advanced Encryption Standard (AES) may be identified to convert the unrealized datasets into unreadable format through encryption. Sample datasets are taken and stored for reuse also and recover the utility of original datasets data reconstruction is performed using. The proposed Fuzzy Decision Tree (FDT) algorithm is identified to reconstructs the original dataset which can be viewed for future use even if the samples lost.