A novel data transformation technique is proposed in order to preserve sensitive information in the datasets. The proposed technique achieved high degree of data distortion and maintained similar accuracy before and after transformation for all the considered data mining algorithms. The performance of the proposed technique is evaluated in terms of data utility, privacy measures, and time and memory requirements using six benchmark datasets. Results indicate good performance when the method is evaluated using popular classifiers. In future, improved methods for transforming nominal attribute values may be explored.
