An efficient K-anonymization algorithm combining C-modes with MDAV

Individual privacy preservation has recently become an increasingly important issue when publishing microdata for mining purpose. K-anonymity is a popular model for protecting privacy, which requires that each record in the released dataset be indistinguishable with at least (k-1) other records with respect to quasi-identifier. MDAV, an efficient k-anonymization algorithm, has been extensively investigated and applied. However MDAVpsilas efficiency decreases dramatically with dataset size increasing. C-modes is an efficient clustering algorithm for large dataset, but which cannot realize k-anonymity. Combining C-modes with MDAV, we propose an efficient algorithm for large dataset k-anonymization problems. Experiments show that, compared with MDAV algorithm, the proposed algorithm increases efficiency dramatically especially for large dataset.