立方体实体化的遗传算法设计与实现

多维立方体由基本立方体和基本立方体的聚集产生的立方体组成.为了实时响应查询,查询优化非常重要.多维立方体实体化是一种很好的解决方案.在整个立方体集中实体化一个适当的子集,满足较好的空间和时间的平衡,并且满足对视图查询频率不同的要求.为了实体化一个合适的立方体子集,通过介绍启发策略,提出了立方体实体化的启发式遗传算法.该算法可以在用户要求的限定的存储空间约束条件下,较快的求出全局最优解.经实验测试,该算法具有良好的性能和快速的求解能力.

Design and implementation of materialized cubes genetic algorithm

Multidimensional data cubes are composed of base cube and other cubes aggregating on base cube. Since response time should be small, query optimization is critical. Materializing multidimensional cubes is a good solution. Partial materialization gives some good tradeoffs between the storage space used and the average time to answer a query. In order to materialize an appropriate cube subset, a heuristic genetic algorithm was proposed. The algorithm can converge quickly and has the ability of optimizing globally in the condition of the given space. It can also satisfy the requirement of different query frequency of views. The experimental studies indicate that the genetic algorithm have good performance.