Scalable Multidimensional Anonymization Algorithm over Big Data Using Map Reduce on Public Cloud

It appears that everybody observes with special attention, the occurrence of big data and its practice. There is no disbelief that the big data uprising has instigated. Though the practices of big data propose favorable business paybacks, there are substantial privacy implications. Multidimensional generalization anonymization scheme is an actual method for data privacy preservation. Top-Down Specialization (TDS) and Bottom-Up generalization (BUG) are two methods to attain multidimensional anonymization. However, prevailing methodologies for multidimensional generalization anonymization scheme disconcerts parallelization proficiency, thereby missing scalability while managing big data on cloud. TDS and BUG suffer from poor performance for certain value of k-anonymity parameter if they are utilized individually. In this paper, we recommend a hybrid method that combines TDS and BUG together for competent multidimensional anonymization over big data. Additionally, Map reduce based algorithms for two components (TDS and BUG) to increase high scalability cloud are designed. Experiment estimations determine that the hybrid method expressively progresses the scalability and proficiency of multidimensional generalization anonymization system over prevailing methods.