Privacy Preserving Data Mining in Distributed System using RDT Framework

Author(s):

Kripa Joy , Mar Athanasius College of Engineering Kothamangalam, Kerala; Aparnasree R, Mar Athanasius College of Engineering Kothamangalam, Kerala; Asst Prof. Linda Sara Mathew, Mar Athanasius College of Engineering Kothamangalam, Kerala

Keywords:

Random Decision Tree (RDT), Distributed System, classification

Abstract:

Distributed data is very important in modern information driven applications. Most of the applications are using distributed databases because of the availability of data in different databases. Use of distributed data is very challenging because of the difficulty of merging the data which are more private. Without losing the privacy of data each application need to maximize the utility of the collected information. Using only local data will not give an optimal utility of the data. In this case techniques for privacy-preserving knowledge discovery is very important. Existing techniques for privacy-preserving data mining are cryptography based techniques and perturbation based technique. Cryptographic techniques are too slow for large scale data sets. Perturbation based technique doesn’t give a much privacy for the data which are distributed. Random Decision Tree framework can used for privacy preserving data mining. Random decision trees (RDT) shows that it is possible to generate equivalent and accurate models with much smaller cost and it is very suitable for parallel and fully distributed architecture.

Other Details:

Manuscript Id	:	IJSTEV3I1145
Published in	:	Volume : 3, Issue : 1
Publication Date	:	01/08/2016
Page(s)	:	329-335

Download Article