面向相似数据集的关系数据库压缩  

Relational Database Compression for Similar Data Sets

作　　者：邓文平[1] 朱培栋[1] 卢锡城[1] 

机构地区：[1]国防科技大学计算机学院,长沙410073

出　　处：《计算机工程》2007年第14期69-70,85,共3页Computer Engineering

基　　金：国家"863"计划基金资助项目(2005AA121570);国家"973"计划基金资助项目(2005CB321801)

摘　　要：采用关系数据库模型进行建模,对于同一关系框架上的数据定义了相似数据集。对单个数据集,通过关系拆分对数据库模型进行规范化处理,去除了关系内部的数据冗余;对多个数据集之间的压缩提出了一种基于0-1状态标记序列的增量式无损压缩算法,压缩后的数据可以快速地完全解压缩。试验结果表明,算法可以实现对相似数据集的高效无损压缩和快速查询。A new definition of similar data set is proposed for some special data sets which have the same attributes in a relational database model. For compression of a single data set, the database normalization is performed by partitioning the relations; for multiplet similar data sets, a data lossless compression algorithm is proposed, which is based on a O-1 status tag sequence. With the compression method, redundancies among similar data sets evidently decrease, and decompression finishe fast and completely. Experimental results show that with the method compression on similar data sets is efficient without any loss and access to the data is fast.

关 键 词：数据库压缩 冗余度 相似数据集 无损压缩 压缩比 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]