International Journal for Development of Computer Science & Technology (IJDCST)
As generalization and bucketization, several anonymization techniques have been designed for publishing privacy preserving micro data. In existing approaches there is some amount of information losses by using generalization particularly in high dimensional data. There is no a clear separation between quasi-identifying attributes and sensitive attributes in case of bucketization. To overcome this problem, the authors proposed an approach called slicing. The data partitions both horizontally and vertically in slicing. Slicing provide better membership disclosure and better data utility over than generalization.