تکنیک های بهینه سازی پرس و جو برای پایگاه داده XML

ABSTRACT Query Optimization Techniques for XML Databases

XML (eXtensible Markup Language) was primarily used as a human consumable exchange format. However, the amount of exchanged XML data often grows exponentially via the Web medium. Web application such as search engine, ecommerce, e-learning portals require advanced tools for managing the data. Communities no longer use the search engine just to retrieve the full-text queries but also request for more specific data (structure queries). This drives the requirement for storing and querying large-scale XML data as efficient and reliable as possible. Several XML query languages have been proposed such as Lorel [1], Quilt [2], XML-GL [3], XPath [4], XQuery [5], XOM [6], XAL [7] and YATL [8]. These query languages utilize regular path expression, thus using the conventional approach such as tree traversal may have performance degrade especially on concurrent access.

This paper is mainly divided into three sections: XML data model, XML storage, indexing techniques and labeling techniques. Section II presents the XML structure and data model. Section III describes the possible alternatives to store XML. Being a semistructured data, there are three possible ways. Section IV presents the optimization techniques via indexing and labeling scheme techniques respectively. Section V wrap-ups and suggest hybrid optimization techniques by comparing the advantages and disadvantages of these techniques. However, the mathematical and theoretical proofs are omitted in this paper.