A Big Data Security using Data Masking Methods

Abstract

Due to Internet of things and social media platforms, raw data is getting generated from systems around us in three sixty degree with respect to time, volume and type. Social networking is increasing rapidly to exploit business advertisements as business demands. In this regard there are many challenges for data management service providers, security is one among them. Data management service providers need to ensure security for their privileged customers in providing accurate and valid data. Since underlying transactional data have varying data characteristics such huge volume, variety and complexity, there is an essence of deploying such data sets on to the big data platforms which can handle structured, semi-structured and un-structured data sets. In this regard we propose a data masking technique for big data security. Data masking ensures proxy of original dataset with a different dataset which is not real but looks realistic. The given data set is masked using modulus operator and the concept of keys. Our experiment advocates enhanced modulus based data masking is better with respect to execution time and space utilization for larger data sets when compared to modulus based data masking. This work will help big data developers, quality analysts in the business domains and provides confidence for end-users in providing data security.