Skyline Computation with Post Filter on Hadoop Framework with Multiple Reducers

Abstract

Today’s world generating huge data that shows exponential growth of data space along with multi-dimensional complexity, with this data query processing becomes difficult. Till now we are using traditional skyline to process all types of data sets. Skyline process demands more computational memory and efficiency for their traditional models. Skyline query processing suffering with the major problems due to increasing in time and space complexity over imbalanced datasets. There is chance of increasing duplicate data sets is another limitation for using traditional skyline processing models to handle spatial uncertain data. Parallel Skyline computation is one solution to handle large spatial data. Parallel skyline computational models are already available, but they also suffering with producing duplicate data sets. By keeping this in mind another solution is proposed that is filtering the data to identify and remove duplicate sets. This idea works and improves query processing speed by eliminating sparsity or empty patterns. In this paper we used k-nearest neighbor approach to eliminate empty pattern or sparsity. If we reduce empty pattern and sparsity it helps to reduce query processing time.