A Protection Layer over MapReduce Framework for Big Data Privacy

Abstract

In many organizations, big data analytics has become a trend in gathering valuable data insights. The framework MapReduce, which is generally used for this purpose, has been accepted by most organizations for its exceptional characteristics. However, because of the availability of significant processing resources, dispersed privacy-sensitive details can be collected quickly, increasing the widespread privacy concerns.  This article reviews some of the existing research articles on the MapReduce framework's privacy issues and proposes an additional layer of privacy protection over the adopted framework. The data is split into bits and processed in the clouds, and two other steps are taken. Hadoop splits the file into bits of a smaller scale. The task tracker then allocates these bits to several mappers. First, the data is split up into key-value pairs, and the intermediate data sets are generated.  The efficiency of the suggested approach may then be effectively interpreted. Overall, the proposed method provides improved scalability. The following figures compare execution time with relation to file size and the number of partitions. As privacy protection technique is used, the loss of data content can be appropriately handled.  It has been demonstrated that MRPL outperforms current methods in terms of CPU optimization, memory usage, and reduced information loss.  Research reveals that the suggested strategy creates significant advantages for Big Data by enhancing privacy and protection. MRPL can considerably solve the privacy issues in Big Data.