基于位图连接索引的复杂多维层次的连接和聚集算法  

A Join and Aggregation Algorithm for Complex Multi-dimensional Hierarchies Based on Bitmap Join Index

作　　者：荀亚玲[1] 张素兰[1] 

机构地区：[1]太原科技大学计算机学院,太原030024

出　　处：《通讯和计算机（中英文版）》2007年第2期10-14,37,共6页Journal of Communication and Computer

摘　　要：在ROLAP中往往涉及到大量数据的复杂即席查询，从SQL角度看，这些查询通常都包含多表连接和分组聚集操作。本文提出了一种连接和聚集操作的新算法JAMDHBJI，该算法充分考虑了ROLAP中复杂多维层次的特点，同时考虑到并非全部维都具有维层次的语义特性，将维层次编码和位图连接索引有效结合，把复杂的连接和分组聚集操作转化为在事实表上的区域查询，从而大大提高了连接和分组聚集的效率。理论分析表明该算法是高效的。ROLAP involves in complex extempore query on a mass of data. These queries include multi-table join and aggregation operation from SQL＇s point of view. As a result, a novel algorithm JAMDHBJI （Join and Aggregation of Multi-dimensional Hierarchies based on Bitmap Join Index） is proposed in this paper. This algorithm not only takes the characteristics of the complex multi-dimensional hierarchies into consideration but also considers some dimensions haven＇t the characteristics and efficiently integrates bitmap join index into dimension hierarchical encoding. It improves the efficiency of join and aggregation by translating join and aggregation operation of complex multi-dimensional hierarchies into range queries of fact table. The analysis shows the performance of JAMDHBJI algorithm has high efficiency.

关 键 词：ROLAP 多表连接 聚集查询 维层次编码 位图连接索引 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]