Query Transformation for Exact Cardinality Computing
In general, query optimizers rely on a cost model to choose an appropriate query execution plan for the given queries. An important key parameter of the cost estimation is the cardinality of sub-expressions of the queries. Traditionally, the optimizers may use the estimation cardinality techniques, which can lead to the estimation errors, and hence the poor execution plans. The exact cardinality approach can be applied to resolve such problem, though its computational expense can be costly. A possible way to improve the efficiency is the query transformation since it can provide the alternation to the optimizers. In this paper, we focus on investigation at the effects of the query transformation to the exact cardinality computing processes. The query transformation techniques to be considered in our work are the traditional but widely applied techniques, i.e. subquery unnesting, group-by view merging, join factorization, and join predicate pushdown. The experiment results on the real-life datasets have been presented to validate such proposed work.
Prariwat Kwanglat Juggapong Natwichai
Computer Engineering Department Faculty of Engineering,Chiang Mai University,Thailand
国际会议
上海
英文
514-518
2011-03-11（万方平台首次上网日期，不代表论文的发表时间）