Optimizing Twig Query Pattern Based On XML Schema

With the rapid development of information technology, XML data is used moreand more widely. XML is applied in many aspects of Web development, commonlyused to simplify data storage and sharing. In order to promote the standardization ofXML data query and processing, the World Wide Web Consortium develops theXQuery language for describing XML data query. XQuery language uses XPathexpression to locate data and its function for XML is similar to SQL for the relationdatabase. Due to the characteristics of semi-structured of XML data, the expense ofXML data query is large. So the study of optimization techniques of XML data query,especially the optimization techniques of the core operation for XML dataquery tree pattern query causes the attention of many scholars.Twig query is also called as tree pattern query. XQuery query put forward byusers often contains redundant query. The containment relationships between thesubqueries may simplify the tree pattern query in order to reduce the amount ofcomputation of tree pattern query and improve the query speed of XML data. Becausea lot of XML datas use XML Schema to describe the data model, many tree querypatterns can be simplified according to specific data model information. This paperanalyes the current situation and shortcomings of the existing methods of queryoptimization, and focuses on the optimization methods of tree pattern which haslogical node AND and OR, proposes an effective method of optimizating tree patternwhich has logical node AND and OR based on the structural constraint relationshipsRPC(Required Parent-Child) and RAD(Required Ancestor-Descendant) extractedfrom XML Schema, deleting the unnecessary branches in the tree pattern. Because thetree pattern having logical node AND and OR preferably satisfies the requirements ofXQuery language to describe the data query, the query optimization method in thispaper supports the high-performance implementation of XQuery better. The paperincludes the design and realization of algorithm of extracting RPC and RADrelationships from XML Schema and algorithm of optimizating the query nodes andlogical nodes in tree pattern based on RPC and RAD relationships. The realizationand test of the algorithms above illustrate that this kind of optimization method iseffective and practical.