Weighted Constraint Satisfaction and Genetic Algorithm to Solve the View Selection Problem

A Data warehouse is a tool that is used by big companies, and it gathers data coming from different sources. The main goal of a data warehouse is not only to store data, but also to help companies to make decisions. The huge volume of data makes processing queries complex and time-consuming. In order to solve this problem, the materialization of views is suggested as a solution to improve the processing of the queries. The materialization of views aims to optimize an objective function which is a compromise between the cost of processing queries and the cost of maintenance under a storage space constraint. In this work, we modelled the view selection problem as a weight constraint satisfaction problem. In addition, we use the used the multiple view processing plans (MVPP) Framework as a search space, and we call genetic algorithm to select views to be materialized. According to experimental result, the proposed algorithm has been used to show the quality of the appropriate materialized views selection.