Volume 06,Issue 02 May 2016 ISSN : 2456- 5083 Page 56 PERTURBATION ORIENTED DATA CONVERSION TECHNIQUE IN BIG DATA WITH PRIVACY PROTECTION * R. SARATH BABU, ** A. NARESH KUMAR * Department of CSE, JNTUH College of Engineering Hyderabad, INDIA ** Department of CSE, JNTUH College of Engineering Hyderabad, INDIA * sharath.rakki@gmail.com, ** amruthamnaresh@gmail.com ABSTRACT: The growing amount of large statistics moreover will increase the danger of breaching the privacy of people. Since large statistics require excessive computational power and a large storage, distributed structures are used. As multiple occasions are concerned in those systems, the chance of privacy violation is extended. There have been some of the privateness- preserving mechanisms developed for privateness protection at particular degrees of a massive data lifecycle. The reason of this paper is to provide a whole assessment of the privateness protection mechanisms in massive data and present the demanding situations for gift mechanisms. The different upcoming challenge inside the area of large facts is Veracity, due to this the trustworthiness of the records this is how comfortable the information are received, saved, processed and transmitted. Henceforth this Veracity is becoming a few different period within the huge records time. So if we're capable of use cryptography and perturbation then it will benefit protection in addition to very fewer possibilities of losing data after making use of the privateness retaining. Specifically, in this paper, we illustrate the infrastructure of huge records and the modern privateness-preserving mechanisms in each level of the massive information lifecycle. Key Terms: cloud computing, attribute-based encryption, outsourced encryption, user revocation, collusion attack. I. INTRODUCTION Due to cutting-edge technological development, the amount of records generated by means of manner of social networking websites, sensor networks, Internet, healthcare packages, and plenty of different companies, is drastically growing every day. All the huge amount of information generated from distinctive resources in a couple of formats with very excessive pace is referred as large records. Big records have turn out to be a very active studies region for closing couple of years. The facts generation charge is developing so swiftly that it's far becoming incredibly difficult to handle it using traditional strategies or structures. Enormous statistics is a term related to the statistics devices whose length is beyond the ability of regularly applied programming frameworks retaining in thoughts the stop aim to store, oversee and prepare. As of past due in modern day statistics-driven global the big facts handling and exam have became out to