A New Technique for Protecting Sensitive Data and Evaluating Clustering Performance
Data mining researchers and policy makers have need of raw data collected from organizations and business companies for their analysis. Any transmission of data to third parties and the organizations outsourcing the authors' work should satisfy the privacy requirements in order to avoid the disclosure of sensitive information. In order to maintain privacy in databases, the confidential data should be protected in the form of modifying the sensitive data items. In statistical disclosure control, masking methods are used for modifying the confidential data. Most of the perturbative masking techniques existing in the literature are general purpose ones. In this paper, a new perturbative masking technique called as Modified Data Transitive Technique (MDTT) is used for protecting the sensitive numerical attribute(s).