Abstract

An efficient materialized view selection approach for data cube utilizing evolutionary optimization

Author(s): Gang Li

In this paper, we focus on the problem of materialized view selection for data cube, which is an important in the research field of database management. In the data warehouse, multi-dimension data can be represented as a data cube, which is a basic element in data warehouse. Particularly, each sub-cube is corresponding to an aggregation view in a specific the data cube. As the objective of materialized view selection for data cube is to minimize the sum of query cost and maintenance cost, in this paper, we converted data cube materialized view selection problem to an evolutionary multi-objective optimization problem. Afterwards, we propose a materialized view selection algorithm for data cube using evolutionary multi-objective optimization. When the stopping condition is satisfied, output of the proposed algorithm can be utilized as the data cube materialized view selection results. To testify the effectiveness of the proposed algorithm, we conduct experiments to make performance evaluation. Compared with other materialized view selection methods, the proposed algorithm performs better in the evaluation criteria “Time cost”, “Average response time”, and “Maintaining and updating time”.