计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (34): 134-135.DOI: 10.3778/j.issn.1002-8331.2009.34.041

• 数据库、信号与信息处理 • 上一篇    下一篇

函数依赖对商覆盖立方体生成效率的影响

彭湘凯1,陈富强1,2   

  1. 1.广东技术师范学院,广州 510665
    2.华南理工大学 计算机科学与工程学院,广州 510641
  • 收稿日期:2009-06-25 修回日期:2009-08-03 出版日期:2009-12-01 发布日期:2009-12-01
  • 通讯作者: 彭湘凯

Computing cover quotient cube considering functional dependency in base tables

PENG Xiang-kai1,CHEN Fu-qiang1,2   

  1. 1.Guangdong Polytechnic Normal University,Guangzhou 510665,China
    2.School of Computer Science & Engineering,South China University of Technology,Guangzhou 510641,China
  • Received:2009-06-25 Revised:2009-08-03 Online:2009-12-01 Published:2009-12-01
  • Contact: PENG Xiang-kai

摘要: 提出一种新的商覆盖立方体生成算法DDFS。指出在基本表维属性之间可能存在函数依赖;分析了这种函数依赖对DFS算法的影响;按照决定子在前的原则调整了DFS算法对基本表进行水平分割时所依据的维属性的顺序;对于存在函数依赖的维属性对,节省了判断是否存在类函数依赖的操作。采用weather数据集进行的实验结果表明,采用DDFS计算商覆盖立方体所需时间仅为采用DFS算法时的76%。得出DDFS较之DFS可以缩短商覆盖立方体生成时间。

关键词: 数据仓库, 商覆盖立方体, 函数依赖, 深度优先探索

Abstract: This paper proposes a new algorithm DDFS to compute cover quotient cube.Functional dependency between base attributes are considered to improve the efficiency of computing.When the base table is partitioned with a particular attribute order,the attribute functionally depended is considered with a priority.For the attributes pair where there is a functional dependency,the cost of finding the congeneric functional dependency is retrenched.Experimental result on the dataset weather shows that the computing time associated with DDFS is only 76% of that with DFS.There is a conclusion that DDFS can shorten the time of computing cover quotient cube,compared with DFS.

Key words: data warehouse, cover quotient cube, functional dependency, Deep First Search(DFS)

中图分类号: