PRIVACY PRESERVING USING DATA ANONYMIZATION TECHNIQUE ON HADOOP

ABSTRACT At present applications are developed widely and also increase various internet based services and cloud applications, the need of cloud environment with buildup facilities is increasing with very widely. Due to the increase in multiple users. A large number of cloud services require users to share private data like hospital health records for data scrutiny or mining, bringing privacy concerns. At present, the scale of data in many cloud applications increases hugely in accordance with the Big Data trend, thereby making it a challenge for commonly used software tools to manage, process, and capture such large-scale data within a tolerable elapsed time. It is challenging for previous annonymization approaches to acquire privacy on large scale data sets due to insufficiency.To preserve privacyamong the data being shared an adequate anonymization technique is used. In the proposed system we can apply data partition MapReduce method.Anonymization the data is important because in some cases,there is no way to stop this information from being available and some data has high value and analyzing it could lead to progress in fields that are highly important, like hospital and education related data. A scalable two-phase top-down specialization (TDS) approach uses MapReduce architecture on cloud to annonymized large scale datasets to design a group of innovative MapReduce jobs to particularly accomplish specialization computation in a highly scalable way. Anonymizing data sets via generalization to satisfy privacy requirements such as k anonymity will be use category of privacy preserving techniques. So the ability of TDS and efficiency of TDS can be significantly upgraded over existing approaches. We can solve scalability problem of large-scale data anonymization by TDS, and can proposed a two-phase TDS approach using MapReduce on cloud. In this paper we are focuses on providing authorized data and preserved identity of authorized data also protect the privacy ofindividuals represented in the data.. In TDS approach data sets can partitioned and anonymized in parallel in the first phase, and can generate midway result.Then,we can merged midway result and can further anonymized to yield consistent kanonymous data sets in the second phase. Experimental evaluation results demonstrate that with our approach, the scalability and efficiency of TDS can be significantly improved over existing approaches