Privacy preserving method for knowledge discovered by data mining
Sara Tedmori · International Journal of Information and Communication Technology · 2018
In spite of its success in a wide variety of applications, data mining technology raises a variety of ethical concerns which include among others privacy, intellectual property rights, and data security. In this paper, the author focuses on the privacy problem of unauthorised use of information obtained from knowledge discovered by secondary usage of data in clustering analysis. To address this problem, the author proposes the use of a combination of isometric data transformation methods as an approach to guarantee that data mining does not breach privacy. The three transformation methods of reflection, rotation, and translation are used to distort confidential numerical attributes for the purposes of satisfying the privacy requirements, while maintaining the general features of the cluster in clustering analysis. Experimental results show that the proposed algorithm is effective and provide acceptable values for balancing privacy and accuracy.