面向大数据的物化视图选择算法  

Materialized view selection algorithms for big data

作　　者：付岩[1,2] 冯径[1] 钱越英 

机构地区：[1]解放军理工大学气象海洋学院 [2]96631部队 [3]北京应用气象研究所

出　　处：《计算机应用》2017年第A01期250-254,共5页journal of Computer Applications

基　　金：国家自然科学基金资助项目(61371119);江苏省自然科学基金资助项目(BK20130070)

摘　　要：使用适当的算法选取视图集进行物化,可以使数据库取得更高的查询效率。从经典的物化视图模型出发,分析比较了不同物化视图选择算法的优缺点,指出传统算法难以解决大数据背景下海量高维度数据的物化视图选择问题,归纳了在云计算并行环境下解决物化视图选择问题研究的进展,提出了支持数据立方体物化视图选择算法的改进思路。It makes the database get higher query processing efficiency when appropriate algorithms are applied to select the materialized view set. In this work, representative frameworks of materialized views were introduced; advantages and disadvantages of different algorithms for materialized view selection were compared. It is pointed out that the conventional algorithms are hard to solve the materialized view selection problem when the data sets are large in amount or the dimensions of them are very high. The latest researches on solving the materialized view selection problem by using parallel computing architecture under cloud computing environment were summarized. Some thoughts were proposed to improve the algorithms for solving materialized view selection problems.

关 键 词：物化视图 选择算法 数据立方体 大数据 云计算 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]