Data Anonymization of Vertically Partitioned Data Using Mapreduce on Cloud

In the world of computers, cloud services, on large scale, are being offered by service providers. User wishes to share some private information that has been stored on the cloud server due to various reasons such as data analysis, data mining and so on. These things bring up a concern about privacy. Privacy preservation may be attained by Anonymization data sets via normalization for satisfying privacy needs by making use of k-anonymity method that happens to be one widely employed kind among the privacy preserving methods. During the current period, data on cloud applications have greatly been found to increase their scale progressively in connection with the trend of Big Data. Therefore it becomes really difficult to manage, accept, process and maintain such huge volume of data within the stipulated time stamps. Hence, it is a very tough task preserving privacy of sensitive and huge sized data. Privacy preservation using existing anonymization methods may not prove efficient because they are not able to handle the scaled datasets. The approach handles anonymization issue on very huge scale cloud datasets by making use of two phase top down specialization method and MapReduce framework. Novel MapReduce tasks are cautiously devised in both the stages of this method for achieving specialization calculation on datasets that are scalable. Efficiency and scalability of the Top down Specialization (TDS) is found to increase significantly over the presently existing method.