Abstract: Data mining emphases on producing beneficial information from the data sources rather than a simple data mining technology. Among several forms of data mining tasks, data classification is a key task for a user who wants to group a record at hand based on the database existing in the cloud. The existing privacy methods for privacy preservation in data mining are perturbation method and secure multiparty method (SMC). But these methods are current for the data that are not encrypted. Hence it is essential to present new protocols for the classification problem in data mining for the database stowed in cloud in encrypted form. We proposed the input record query classification problem of data mining task for accessing the database in encrypted form in the cloud is solved using KNN Classification method. A new privacy preserving protocol based on KNN classification method is used for protecting the privacy preservation and confidentiality of the data in the database, input query and data access.
Keywords: Security, KNN classification, Outsourced databases, Encryption.