数据仓库中一种有效的高维联机分析处理方法

A improved parallel processing approach for high-dimensional OLAP in data warehouse

在大量数据仓库系统中,对于一个d维的data cube,数据立方体(cube)可以生成2d个聚集cuboids,然而随着数据仓库维数的增长,计算这些预聚集数据已经成为一个瓶颈.在minimal cubing方法的基础上,提出一种具体层次语义特性的多维层次数据立方体——前缀索引立方体(prefix-index cubing)技术,将高维cube划分成若干个低维立方体cube,以实现高维cube的分布式存储和并行计算.理论分析与实验结果表明,相对于以往的minimal cubing等方法,前缀索引立方体方法的性能显著提高.

In many data warehouses, it can generate 2d cuboids for the cube with d dimensions. However, as the size of data warehouses grows, the time it takes to perform this pre-computation becomes a significant performance bottleneck. In this paper, a multi-dimensional hierarchical cubing approach, called prefix-index cubing approach, is proposed based on an extension of the previous minimal cubing approach. This method partitions the high dimensional data cube into low dimensional cube segments. The proposed data allocation and processing model support distributed storage and parallel processing, as well as load balancing for disks and processors. The analytical and experimental results show that the proposed method is significantly more efficient than other existing cubing methods such as minimal cubing approach.