Selection of Optimal Materialized Views in Data Warehouse Using Hybrid Technique

Decision making is the main purpose of DW. Typically, decision making queries are analytical, complex, recurring and include aggregation functions or many join operations posed over DW. A critical issue in designing DW is answering these queries efficiently. Many ways have been proposed to address this problem, one of them is materialize views while due to space constraints all views cannot be materialized in DW. In this paper we have design an efficient methodology for selecting an optimal MVs based on three factors (MV response time, MV storage area and MV frequency) using bitmap index to minimize the total time of creating MVs, then using hybrid technique (Firefly algorithm and Quantum Particle Swarm Optimization algorithm) to select optimal MV which has low response time, low storage area and high frequency. The results proved that bitmap index achieved good results because it take less time and storage area in recoupoing results since bitmap indices have the ability of accomplishing processes on index level before recouping the base relations where, the total time of applying these functions over bitmap index was found 168 milleseconds, while directly over base tables was found 216 milleseconds. Also, the hybrid technique was more efficient in term of optimal MVs selection time since, FA presents optimal initial point to QPSO algorithm.