Return to author index

Volume: 3 Issue: 4

Selection of views to materialize in data warehouse: A hybrid solution

A. Boukra, M. Ahmed Nacer, S. Bouroubi

Abstract:
The idea of a data warehouse was introduced so as to make decisions efficiently. One of the determining parameters in this efficiency is the queries response time. One of the techniques used to reduce this delay is to select a set of views and to materialize them. The occurring problem is to find the optimal set of views that minimizes the total response time, by respecting some resources constraints. The exact algorithms of operational research give good results only for small size problem. For more important sizes, heuristics are used to give solutions close to the optimal one. In this paper, we propose a new algorithm, based on the evolutionary algorithms which replace the crossing and mutation operators by an algorithm based on the ant colony. Proceeding this way, we eliminate the blind aspect from these two operators. The experimental results show that the performance ratio between our algorithm and the optimal algorithm is 0.90 [1] and generates a satisfying number of feasible solutions.

Keywords:
OLAP, Data Warehouse, Materialized views, Dimensional modeling, Evolutionary algorithm, Ant colony Algorithm.

doi:10.5019/j.ijcir.2004.113

Full Text PDF
















^ TOP