摘要
基本单元组浓缩把那些由同一条基表元组聚集计算得到的立方元组浓缩成一条,从而减小数据立方的体积。共享前缀技术通过消除元组之间的前缀冗余来进一步压缩数据立方。PrefixCube就是将基本单元组的浓缩与共享前缀结合,而提出的一种有效的数据立方组织形式。在本文中,我们以批处理模式进一步优化计算PrefixCube,从而减少PrefixCube的计算时间代价。通过在模拟数据集和真实数据集上的实验证明,在大多数数据集上,以批处理模式计算PrefixCube要优于一般模式计算PrefixCube。
BST Condensing is an effective approach to reducing cube size, which condenses those tuples, aggregated from the same single base relation tuple, into one physical tuple. Prefix-sharng technique can further reduce the size of a data cube, by eliminating prefix redundancies existing among cube tuples. PrefixCube is proposed to be an efficient cube structure by augmenting BST condensing and prefix-sharing. In this paper, we optimize the computation of PrefixCube through batch mode processing to reduce the computation time cost. Through extensive experiments, using both synthetic and real world dataset, the batch-mode computation of PrefixCube is proved to outperform the normal-mode computation on most of datasets.
出处
《计算机科学》
CSCD
北大核心
2004年第12期81-85,96,共6页
Computer Science
基金
国家自然科学基金(项目编号60303030)
