SENSITIVE DATA MONITORING IN PARTIONING TECHNIQUE

Numerous anonymization methods, such as simplification and bucketing, are designed for privacy preserving micro data publishing. Many works shown that generalization lose considerable amount of information, especially for high dimensional data object. But Bucketization does not prevent membership disclosures and do not apply for data that do not have a clear separation among quasi-identifier characteristics and delicate attributes. In this work, we present a novel technique called slicing partitioning on the data both horizontally and vertically. We show that slicing preserves data utility better than oversimplification and can be castoff for involvement disclosure protections. Another important use of slicing is that it can be used in high-dimensional data. We show how slicing can be used in attribute disclosure protection and developed an effectual procedure for adding the sliced data that obey diversity requirements. Our experiment confirms that slicing preserves better utility than generalizations and more effective than bucketizations in workloads involving the sensitive attributes. Our work also demonstrates that slicing can be used to prevent membership disclosure