Shearing based data transformation approach for privacy preserving clustering

Data mining is a technique used to extract un-known patterns from large volume of data. Privacy has be-come a major concern in data mining. The data which are stored in the organization´s database may have some confi-dential information and hence it has to be protected from any unauthorized usage. In order to overcome the privacy problem in this paper we propose a new technique called shearing based composite transformation. Data owner trans-forms the original data into distorted data by shearing based composite data transformation. Only this distorted data is given to the clients. For clustering we have used k-means algorithm and from our experiments we found that the total number of elements in the clusters is same with the original and distorted data.