数据立方体聚集范围查询分块方法研究  

The Research on the Partition for Aggregation Range Queries in Data Cube

作　　者：师智斌[1] 黄厚宽[1] 

机构地区：[1]北京交通大学计算机与信息技术学院,北京100044

出　　处：《计算机科学》2007年第12期93-96,共4页Computer Science

摘　　要：范围查询是数据立方体数据分析的有效工具,预计算技术通过预先计算并存储范围查询的结果,可以实现快速的用户响应。近年来研究人员对基于MOLAP的预计算技术的研究主要以prefixsum及分块技术为基础。本文对预计算技术的分块方法进行研究,分析了现有分块技术的方法和性能,并提出了两种新的分块方法:嵌套分块和基于前缀区域边界的分块。本文对这两种分块的方法和特点做了阐述,研究表明这两种方法为分块技术提出了新的思路,是对现有分块方案的有力补充。A range sum query is one of effective tools to analyze data in data cubes. Pre-computing can speed response times of range query through computing and storing the query result on-the-fly. The researches on pre-computing based on MOLAP are mostly based on technologies of prefix sum and partition recently. This paper works on partition scheme and analyzes the methods and capabilities of current technologies of partition. Two new methods of partition are put forward in this paper. They are nesting partition and partition based on the border of prefix region respectively. This paper discusses constitution and characteristic of two methods and it shows that they bring forward the new means to partition in data cube and supplement the current methods of partition.

关 键 词：数据立方体 范围查询 划分 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]