期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2015
卷号：4
期号：8
页码：3378-3387
出版社：Shri Pannalal Research Institute of Technolgy
摘要：Partitioning is a design technique applied to thedata warehouse schema for dividing a single table into two ormore partitions (fragments) for improving query performance.Referential partitioning technique can be applied in a datawarehouse, where a fact table is partitioned with reference tothe partitions of the dimension table. The dimension tableselection is an important factor in order to referentiallypartition a fact table. Depending on the number of dimensionand its attributes the possible fragments might be very large.As the data warehouse administrator can manage only a limitedset of fragments in the underlying database it is essential toselect only optimal set of fragments during the design, whichcan reduce the overall query execution cost. In this paper wehave proposed an optimized approach for horizontal referentialpartitioning of the data warehouse schema. We have evaluatedthe proposed approach using TPC-H (SSB) benchmark. Fromthe obtained results it has been observed that the proposedapproach improves the query performance compared to theexisting works.
