高级检索

    一种多维数据库和多维视图模型

    A MULTIDIMENSIONAL DATABASE AND VIEW MODEL

    • 摘要: “多维”是数据仓库/OLAP应用普遍采用的模型化方法,文中提出了一种对多维数据库和在多维数据库上进行的查询/统计进行模型化的方法,并给出了模型的形式化定义.鉴于数据库本身与一次具体任务产生的结果两种对象的性质的不同,特别区分了多维数据库和多维视图.多维数据库是对全部数据及其内在关系的全面描述;多维视图则是针对一次任务的对多维数据库的选择和重组.文中还着重讨论了多维数据库维的划分问题,以函数依赖为主要依据,提出各维应满足正交限制条件,以免出现稀疏矩阵异常,并给出了维的划分算法

       

      Abstract: ultidimensional” is a widely adopted modeling method in the field of data warehouse/OLAP. Here proposed are a formal definition of a user oriented multidimensional database and view model, and a method for deriving that model. Realizing that the multidimensional database is essentially different from multidimensional view,a distinction is made between these two concepts. Multidimensional database is the full description of the whole data and relations among them, while multidimensional view is just a selection and reconstruction of a multidimensional database aiming at satisfying one query task. Based on function reliance,it is pointed out that dimensions must be orthogonal from each other to avoid sparse matrix abnormality. An algorithm for dividing dimensions is also given.

       

    /

    返回文章
    返回