Enhancement of Privacy Preserving Technique using Slicing with Entity Resolution

The various anonymization techniques, called generalization and bucketization, have been designed for providing data privacy and preserving micro data publishing. Recent work shows that generalization loses considerable amount of information on high dimensional data and bucketization did not prevent membership disclosure with clear separation between quasi-identifying attributes and sensitive attributes. The slicing techniques partitions the data both horizontally and vertically is proposed with entity resolution which preserves data utility and membership disclosure protection .Slicing develops an efficient algorithm with the ‘l-diversity requirement. The workload experiments with sensitive attribute ensures that slicing provides better utility than generalization and is more effective than bucketization. As an extension we proposed a technique called overlapped slicing, were the attributes are divided into more than one column and release in each column consists of more attribute correlations.