基于划分的数据仓库查询归并缓冲算法  

MERGING AND BUFFERING ALGORITHM INQUIRY IN PARTITION BASED DATA WAREHOUSE QUERY TECHNOLOGY

作　　者：冯建华[1] 张季[1] 周立柱[1] 邓敏[1] 

机构地区：[1]清华大学计算机科学与技术系,北京100084

出　　处：《计算机应用》2001年第9期8-11,共4页journal of Computer Applications

基　　金：国家"973"重点基础研究发展项目 (G1 9980 30 4 1 4 )

摘　　要：在数据仓库中存在着大量的数据。联机分析处理包含着对大量数据的复杂的查询过程。在对这些数据的存储与查询中都遇到了许多困难。解决这一问题的有效办法就是先将数据划分成便于处理的数据块 ,再分别对每个数据块进行处理 ,最后将各个数据块的处理结果归并在一起。对几种常用的归并算法进行了比较 。Data warehouses contain large amounts of information and on-line analytical processing(OLAP), involve hundreds of complex aggregate queries over large volumes of data. In the process of storing and querying on these data, we are faced with many difficulties. One effective access to solve this problem is to partition these data into data blocks, which are easy to process and process these data blocks individually. Then merge the results together. In this paper, We compared several typical merging algorithms, and discussed the assignment of data buffer in the process of merging.

关 键 词：磁盘缓冲 数据划分 数据仓库 数据库 数据查询归并缓冲算法 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]