In the context of XML data management systems, the estimation of query cardinality is becoming more and more important: the information provided by a query result estimator can be used as input to the query optimizer, as an early feedback to user queries, as well as input for determining an optimal storage schema, and it may be helpful in embedded query execution. Existing estimation models for XML queries focus on particular aspects of XML querying, such as the estimation of path and twig expression cardinality, and they do not deal with the problem of predicting the cardinality of general XQuery queries. This paper presents a framework for estimating XML query cardinality. The framework provides facilities for estimating result size of FLWR queries, hence allowing the model designer to concentrate her efforts on the development of adequate and accurate, while concise, statistic summaries for XML data. The framework can also be used for extending existing models to a wider class of XML queries.
A General Framework for Estimating XML Query Cardinality
SARTIANI, CARLO
2003-01-01
Abstract
In the context of XML data management systems, the estimation of query cardinality is becoming more and more important: the information provided by a query result estimator can be used as input to the query optimizer, as an early feedback to user queries, as well as input for determining an optimal storage schema, and it may be helpful in embedded query execution. Existing estimation models for XML queries focus on particular aspects of XML querying, such as the estimation of path and twig expression cardinality, and they do not deal with the problem of predicting the cardinality of general XQuery queries. This paper presents a framework for estimating XML query cardinality. The framework provides facilities for estimating result size of FLWR queries, hence allowing the model designer to concentrate her efforts on the development of adequate and accurate, while concise, statistic summaries for XML data. The framework can also be used for extending existing models to a wider class of XML queries.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.