Privacy Preserving Data Mining Using Random Decision Tree Over Partition Data: Survey

College of Science, Knowledge University, Erbil, Kurdistan Region, Iraq

^* Corresponding author: nashwan.adnan92@gmail.com

Abstract

The development of data mining with data protection and data utility can manage distributed data efficiently. This paper revisits the concepts and techniques of privacy-preserving Random Decision Tree (RDT). In existing systems, cryptography-based techniques are effective at managing distributed information. Privacy-preserving RDT handles distributed information efficiently. Privacy-preserving RDT gives better precision data mining while preserving information and reducing the calculation time. This paper deals with this headway in privacy-preserving data mining technology utilizing emphasized approach of RDT. RDT gives preferable productivity and information privacy than cryptographic technique. Various data mining tasks utilize RDT, like classification, relapse, ranking, and different classifications. Privacy-preserving RDT utilizes both randomization and the cryptographic method, giving information privacy for some decision tree-based learning tasks; this is an effective technique for data mining with privacy-preserving distributed information. Thus, in horizontal partitioning of the dataset, parties gather information for various entities but have data for all attributes. On the other hand, various associations may gather different data about a similar set of people. Thus, in vertically partitioned data, all parties gather data for the same collection of items. In all of these cases, both horizontal and vertical partitioning of datasets is somewhat inaccurate.