Abstract: - In this paper, a method for estimating the size of relational query results is proposed. The approach is based on the estimates of the attribute distinct values. On the basis of our method, a set of parameters, the so-called Canonical Coefficients, can be derived from actual data; they allow us to approximate both the multivariate data distribution and distinct values of attributes. In particular, the capability of analytic method to estimate selectivity factors of relational operations is considered. Some experimental results on real databases are also presented which show the promising performance of our analytic approach.
Analytic-based estimation of query result sizes
LEFONS, Ezio;TANGORRA, Filippo
2005-01-01
Abstract
Abstract: - In this paper, a method for estimating the size of relational query results is proposed. The approach is based on the estimates of the attribute distinct values. On the basis of our method, a set of parameters, the so-called Canonical Coefficients, can be derived from actual data; they allow us to approximate both the multivariate data distribution and distinct values of attributes. In particular, the capability of analytic method to estimate selectivity factors of relational operations is considered. Some experimental results on real databases are also presented which show the promising performance of our analytic approach.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.