支持非覆盖维的扩展多维数据模型  

An Extended Multidimensional Data Model Supporting Non-Covering Dimensions

作　　者：李泽海[1] 孙吉贵[1] 赵君[1] 于海鸿[1] 

机构地区：[1]吉林大学计算机学院符号计算与知识工程教育部重点实验室,长春130012

出　　处：《计算机科学》2005年第9期113-118,共6页Computer Science

基　　金：国家自然科学基金(批准号:60073039;60273080);吉林省杰出青年基金;吉林省科技发展计划重大项目资助课题

摘　　要：非覆盖维等复杂结构维的处理一直都是数据仓库领域的难题,本文在详细分析非覆盖维特性的基础上,提出了一个扩展的多维数据模型,改变了传统的级别间的映射关系的定义方式,定义了从父级别到子级别分区的映射关系,从而实现了对非覆盖维和非平衡维的支持,并能够完整地表达各种复杂维层次结构语义。同时,基于DAG图描述的维层次结构,定义了基于该多维模型的立方体代数和OLAP操作。将多维数据库概念模型中的维与度量的转换操作引入到OLAP操作集合中,使其支持复杂维的转化操作,进而丰富和增强基于该模型的OLAP系统的分析能力。A new multidimensional conceptual model is proposed here based on current multidimensional models. The mapping from a parent level to one partition of a child level is defined to support irregular relationships among levels, such as non-covering and non-balanced relationships. Thus, the model can express the semantics of non-covering dimensions. Moreover, DAG is used to describe a dimensional hierarchy structure. The relationship between views and cubes is discussed explicitly, and the cube algebra and OLAP operation set are also defined. Especially, the transforming operations between dimensions and measures are also imported to enhance the analytical ability of the model.

关 键 词：数据仓库 联机分析处理 多维数据模型 非覆盖维 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]