Research For XML Query Optimization Technology Based On Relational Database

XML is released as a markup language by W3C in 1998. As a result of its unique technique advantage, XML rapidly becomes the standard of information description and exchange standards on the Internet. With the more and more widely used of the XML, the XML documents has been increased with each passing day, and in order to storage or access the XML data efficiently, the research of XML query optimization based on database technology has become a very important topic.This paper first analyzes the development of XML database, and introduces some basic concepts of XML, such as XML Schema and so on. With the purpose of support XML query optimation better, there is a thorough study about coding of XML data and relational storage model. This paper studies the XML query optimation techniques mainly based on the relational DB, what has done including the following aspects:(1)Based on the region coding and XRel storage model, an extended storage model is proposed for relational database, which can be used to calculate the structure relationalship of the XML doucument efficiently.(2)Relization of a kind of XPth expression optimization which on basis of XML Schema, the implementation steps include:extraction of the Schema information, determine whether the nodes can be optimized or not, and description of the optimization algorithm. Generally speaking, when the XPath expression is much longer, this method can achieve a better optimization by the way of reduce the join operations.(3)As the query of XML document is often converted to the calculation of containion or location relationship between two nodes, to support the structural join algorithms is significant for XML queries. This paper focues on the structural join algorithms, and makes an improved algorithm based on the advanced relational storage model. The improved algorithm in the parent/child relationship, its CPU operating time and I/O complexity is much lower than the original algorithm, especially when the queue of parent nodes exist nested nodes, the improved algorithm is superior.