物化视图选择的一种两阶段优化算法  

A Two-Phase Optimization Algorithm for Materialized View Selection

作　　者：徐海涛[1] 吴昊[1] 

机构地区：[1]杭州电子科技大学软件工程学院,浙江杭州310037

出　　处：《计算机时代》2010年第7期15-18,共4页Computer Era

摘　　要：物化视图是数据仓库中提高查询效率的有效方法,物化视图选择问题是数据仓库设计时期最重要的决定之一。通过研究和实验,提出了一种结合迭代改进算法和模拟退火算法的两阶段优化算法,用于解决物化视图的选择。理论分析和实验结果表明,该算法有效地解决了传统模拟退火算法收敛过慢的缺点,并且其解的质量逼近经典贪婪算法。Materialized view is an effective method for improving the query efficiency in data warehouse,and materialized view selection is one of the most important decisions during the design of data warehouse. Through study and experiment,a two-phase optimization algorithm combining improved iterative algorithm and simulated annealing algorithm is offered to solve the problem of materialized view selection. Theory analysis and experiment result prove that the algorithm has effectively solved the defect of slow convergence in traditional simulated annealing algorithm and the quality of its solution approaches traditional greedy algorithm.

关 键 词：物化视图选择 两阶段优化算法 模拟退火算法 贪婪算法 

分 类 号：TP311.132[自动化与计算机技术—计算机软件与理论]