• 论文 •    

基于数据流模式的聚集快速查询计算研究

陈昕,陈维兴,苏锦祥   

  1. 1.北京信息工程学院计算机信息系统系,北京100101;2.郑州大学计算机科学工程系,河南郑州450052
  • 出版日期:2004-06-15 发布日期:2004-06-25

Aggregate Approximate Query Computation Based on Data Stream

CHEN Xin,CHEN Wei-xing,SU Jin-xiang   

  1. 1.Dep. of Computer Info. System, Beijing Info. Tech. Inst., Beijing100101, China;2. Dep. of Computer Sci. Eng., Zhengzhou Univ., Zhengzhou450052, China
  • Online:2004-06-15 Published:2004-06-25

摘要: 为了减少在计算和存档数据流的过程中对昂贵的存档数据的存取,分析了聚集数据查询的特点,并引入了聚集查询语言和聚集查询重写方法。针对数据流近似查询计算的需求,概述了数据流计算的基本特征,提出了一种基于数据流的近似计算模型。在此基础上,根据实际应用环境给出了一种近似查询提纲的小波方法。该计算架构和方法在人口信息系统中的实际应用表明,可以实现对数据流的快速聚集查询处理,并获得有效的查询结果。

关键词: 集查询计算, 数据流计算, 近似查询方法

Abstract: To reduce the expensive access to archive data, the characteristics of aggregate data query were analyzed, and the aggregate query language and its rewriting method were proposed. Considering the requirement of data stream computation, the basic characteristics of data stream computation were briefly introduced and a kind of framework of approximate computation was put forward based on data stream. Furthermore, a kind of wavelet method was given to form approximately query sketch on data stream according to the practical application environment. This method has been applied to the population statistic information system. It proves that the method can realize rapid aggregate query computation and gain effective outcome.

Key words: aggregate query computation, data stream computation, approximate query method

中图分类号: