A Privacy Preserving K-Nearest Neighbor Classification Protocol Over Secure Encrypted Data

Abstract: Data mining has wide variety of real time application in many fields such as financial, telecommunication, biological, and among government agencies. Classification is the one of the main tasks in data mining. For the past few years, due to the increment in various privacy problems, many conceptual and feasible solutions to the classification problem have been proposed under different certainty prototype. With the increment of distributed computing users have an opportunity to offload the data and processing in the cloud, in an encrypted form. The data in the distributed are in encrypted form, existing privacy preserving classification systems are not relevant. To perform privacy preserving k-NN classification over encrypted data. The recommended protocol preserves the privacy of data, protect the user query, and hide the access mode.

Keywords: Security, K-NN Classifier, Outsourced databases, Encrypted,Privacy Preserving classification.