Graph Based Local Recoding for Data Anonymization

Releasing person specific data could potentially reveal the sensitive information of an individual. kanonymityis an approach for protecting the individual privacy where the data is formed into set ofequivalence classes in which each class share the same values. Among several methods, local recodingbased generalization is an effective method to accomplish k-anonymization. In this paper, we proposed aminimum spanning tree partitioning based approach to achieve local recoding. We achieve it in twophases. During the first phase, MST is constructed using concept hierarchical and the distances amongdata points are considered as the weights of MST and in the next phase we generate the equivalence classesadhering to the anonymity requirement. Experiments show that our proposed local recoding frameworkproduces better quality in published tables than existing Mondrian global recoding and k-memberclustering approaches