Scalable TPTDS Data Anonymization over Cloud using MapReduce

Abstract

With the rapid advancement of big data digital age, large amount data is collected, mined and published. Data publishing become day today routine activity. Cloud computing is best suitable model to support big data applications. Large number of cloud service need users to share microdata like electronic health records, data containing financial transactions so that they can analyze this data. But one of the major issues in moving toward cloud is privacy threats. Data anonymization techniques are widely used to combat with privacy concerns .Anonymizing data sets using generalization to achieve k-anonymity is one of the privacy preserving techniques. Currently, the scale of data in many cloud applications is increasing massively in accordance with the Big Data tendency, thereby making it a difficult for commonly used software tools to capture, handle, manage and process such large-scale datasets. As a result it is challenge for existing approaches for achieving anonymization for large scale data sets due to their inefficiency to support scalability. This paper presents two phase top down specialization approach to anonymize large scale datasets .This approach uses MapReduce framework on cloud, so that it will be highly scalable and efficient. Here we introduce the scheduling mechanism called Optimized Balanced Scheduling to apply the Anonymization. OBS means individual dataset have the separate sensitive field. Every data set consist of sensitive field and give priority for this sensitive field. Then apply Anonymization on this sensitive field only depending upon the scheduling.
DOI: 10.17762/ijritcc2321-8169.150775