首页> 外文学位 >Distributive and algebraic aggregation computation in multidimensional database systems.
【24h】

Distributive and algebraic aggregation computation in multidimensional database systems.

机译:多维数据库系统中的分布式和代数聚合计算。

获取原文
获取原文并翻译 | 示例

摘要

Multidimensional aggregation plays an important role in systems that maintain large data sets. A conceptual Multidimensional Aggregation Object (MAO), which consists of measures, scopes and aggregation functions, is introduced to represent relationships among aggregators for the entire addressable data set.; In the MAO model, aggregations of low-level (intermediate) data can be reused for aggregations on high-level data along the same dimension. Efficient caching of intermediate aggregated data is presented to improve performance. Intermediate data need to be synchronized when fact data are updated, so we propose direct and indirect compensating as well as fully recomputing cache-updating approaches for this purpose.; Both caching and maintenance methodologies can be applied to data indexing if the index transformation is a homomorphism of the original data domain. This implies that systems are able to maintain traditionally independently developed index structures as well as the basic input data.; In order for the system to perform according to our methodologies, a meta-data script language called Execution Plans is presented. Experimental results reveal significant performance improvements when using MAO for distributive and algebraic aggregations.; The proposed data aggregation technique can be applied to equip data-warehousing environments with additional integrated computation capabilities. OLAP systems can benefit from being able to prepare well-maintained calculated results. Our MAO model can enhance data mining tasks by providing necessary aggregations as well as organized data indexing. Decision making and scientific computing systems can all take advantage of our versatile aggregations to compute very large data sets.
机译:多维聚合在维护大型数据集的系统中起着重要作用。引入了一个概念性的多维聚合对象(MAO),它由度量,范围和聚合函数组成,用于表示整个可寻址数据集的聚合器之间的关系。在MAO模型中,低级(中间)数据的聚集可沿相同维度重新用于高级数据的聚集。提出了有效的中间聚合数据缓存,以提高性能。当事实数据更新时,中间数据需要同步,因此我们建议直接和间接补偿以及完全重新计算缓存更新方法。如果索引转换是原始数据域的同构,则缓存和维护方法都可以应用于数据索引。这意味着系统能够维护传统上独立开发的索引结构以及基本输入数据。为了使系统根据我们的方法执行,提出了一种称为执行计划的元数据脚本语言。实验结果表明,使用MAO进行分布和代数聚合时,性能显着提高。所提出的数据聚合技术可以应用于为数据仓库环境配备附加的集成计算功能。 OLAP系统可以从准备良好的计算结果中受益。我们的MAO模型可以通过提供必要的汇总以及有组织的数据索引来增强数据挖掘任务。决策和科学计算系统都可以利用我们的多功能汇总来计算非常大的数据集。

著录项

  • 作者

    Tsai, Meng-Feng.;

  • 作者单位

    University of California, Los Angeles.;

  • 授予单位 University of California, Los Angeles.;
  • 学科 Computer Science.
  • 学位 Ph.D.
  • 年度 2004
  • 页码 100 p.
  • 总页数 100
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 自动化技术、计算机技术;
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号