Research And Implementation For XML Query Optimization Technology Based On Regular Path Expression

With the rapid development of network technology during the recent years, Extensible Markup Language (XML) has become the new criterion of data representation and exchange in Internet. XML, as a kind of semi-structured data, is different from the data models of traditional databases, so it can not be effectively processed by the traditional database techniques. With the development of XML's query language and store mechanism, therefore, how to improve the query efficiency to achieve query optimization has become more and more important.After introducing the current research of XML query and its optimization techniques in details, this paper makes an analysis on the characteristic of regular path expression based XML query techniques, pointing out several problems which need to be improved. Based on finite state automation, this paper presents a new efficient query rewriting method, making use of materialized semi-structured view to rewrite user's query, to improve the query efficiency. An XML query processing prototype system is also designed and realized in this paper. The main research work and improvement are concluded as follows.First, a view index method is presented in the paper. In the new method, the character value of each view in the view set is extracted to build up a view index, with which the efficiency of searching useful views is highly improved.Second, an efficient partial query rewriting method is presented in the paper. Since efficient view can only partially, not fully, rewrite query, the paper presents a new method to replace partial query by views, to fully use the view information to optimize query and improve query efficiency.Third, according to the new query optimization method, an efficient XML query processing prototype system is designed and realized in the paper. For each of user's queries, the system, which is composed of cache management and query processing, will first optimize query rewriting by using the views in the cache, before calculating the query result. Moreover, the paper also presents two scenarios to test the system. It's verified from the experiment results that the new query optimization method presented in this paper can highly improve the XML query efficiency.