Multi-Objective Big Data View Materialization Using Improved Strength Pareto Evolutionary Algorithm

Abstract

Big data view materialization enhances the performance of Big data queries. This is a complex problem due to large volume, heterogeneity, high rate of data generation, low integrity and low value of Big data. Big data view materialization is a bi-objective optimization problem with the objectives - minimization of query evaluation time for a set of workload queries over a window of time and minimization of update processing cost of the views. Structure of Big data views can be represented as directed graph, which can be used to identify the candidate Big data views for a given set of queries. Evolutionary algorithms can be used to solve the problem of Big data view materialization. This paper presents an algorithm based on Strength Pareto Evolutionary Algorithm (SPEA-2) to generate a set of optimal solutions to the bi-objective Big data view selection problem.