摘要
数据切片体现了数据立方体某方面的数据特征,当产生数据切片的数据立方体由于某种原因不可访问时,通过已知的多个数据切片尽可能地恢复数据立方体中的原始信息,有利于对数据的深入分析和理解.提出了一种基于数据切片重建数据立方体的方法,通过数据切片的连接生成多维细粒度空间,利用数据覆盖操作逐步细分每个数据切片所对应的多维空间,以逼近多维细粒度空间.提出了重建后数据立方体的可查询判据.该方法直接利用数据库中的基本操作来实现,高效地支持大数据量的应用环境.
Data slices present some properties of data cube. When data cube cannot be accessed for some reasons, it is very useful to reconstruct data cube based on data slices to discover the implicit information. Therefore, in order to reconstruct data cube, a practical method named Data Cover Method is presented, which creates the multidimensional fine-granularity space by the join operation between data slices, and splits the multidimensional space of data slice into multidimensional fine-granularity space by the data cover operators. An answerable criterion of reconstructed data cube is also presented. This method can be implemented by the basic operations of the DBMS (database management system), and support the application environment of large data set efficiently.
出处
《软件学报》
EI
CSCD
北大核心
2002年第8期1586-1592,共7页
Journal of Software
