Efficient Computing of Skyline Queries using SKY-MR+

Abstract

The skyline operator is introduced due to its wide range of  applications but this process is challenging in case of big data. Mapreduce framework is considered for applications that are data-extensive. For parallel processing of the application SKY-MR+ with the mapreduce is used..In this process the benefit of splitting the terms is done based on estimated execution time. Dominance power ?ltering method  is applied to effectively eliminate non-skyline points . Data partitioning is done based on the region surrounded by the quad tree.It is checked whether each and every skyline candidate point is actually a skyline point using MapReduce.  Workload balancing techniques are used to make estimated execution time of all the machines same. Experiments are done to compare SKY-MR+with the state-of- the-art algorithms using MapReduce.