Efficient systematic clustering method for k -anonymization

摘要: This paper presents a clustering (Clustering partitions record into clusters such that records within cluster are similar to each other, while in different most distinct from one another.) based k-anonymization technique minimize the information loss at same time assuring data quality. Privacy preservation of individuals has drawn considerable interests mining research. The k-anonymity model proposed by Samarati and Sweeney is practical approach for privacy been studied extensively last few years. Anonymization methods via generalization or suppression able protect private information, but lose valued information. challenge how during anonymization process. We refer as systematic problem which analysed this paper. adopts group-similar together then anonymizes group individually. structure defined investigated through paradigm properties. An algorithm developed shown complexity $${O(\frac{n^{2}}{k})}$$, where n total number containing concerning their privacy. Experimental results show our method attains reasonable dominance with respect both execution time. Finally illustrates usability incremental datasets.