Design And Development Of Novel Approach For Privacy Preservation Of Utility Patterns In Knowledge Discovery Process
The huge applications of data mining technologies have raised the concerns about securing information against
unauthorized access, which serves as the important goal of database security and privacy. Privacy and security are the most
significant essentialities, when data is distributed. Privacy is a term that is associated with mining task, with which we are
able to hide some crucial information that we don’t want to disclose to the public. The important consideration in privacy
preservation is to provide a proper balance between privacy protection and knowledge discovery. In order to handle these
scenarios carefully, we propose a privacy preserving utility mining method in this paper based on the process, namely,
sanitization, measure reduction in mining and post-reduction of sensitivity. The data sanitization is the process of identifying
and reducing the sensitive attributes from the database and in the second way, the sensitive attributes are identified through
the knowledge discovery process and the measure is reduced. In the third way, the utility pattern mined from the database is
converted into insensitive utility patterns. The proposed approach is designed to handle privacy protection effectively using
these three ways. Here, the utility pattern mining algorithm is devised utilizing the tree-based data structure and then, the
privacy protection schemes are applied. The performance of the proposed approach is evaluated with the help of benchmark
databases and three different evaluation metrics such as, hiding failure, miss cost and database difference ratio.The proposed
approach is implemented using JAVA.
Keywords— Utility Mining, Sanitization, Privacy Protection, Knowledge Discovery, Database Security.