Optimizing the evaluation of XPath using description logics

Abstract

The growing use of XML in commercial as well as non-commercial domains to transport information poses new challenges to concepts to access this information. Common ways to access parts of a document use XPath-expressions. We provide a transformation of DTDs into a knowledge base in Description Logic. We use reasoning capabilities grounded in description logics to decide if a given XPath can be satisfied by a document, and to guide the search of XML-Processors into possibly successful branches of the document, thus avoiding parts of the document that will not yield results. The extension towards object oriented subclassing schemes opens this approach towards OODB-queries. In contrast to other approaches we do not use any kind of graph representing the document structure, and no steps towards incorporation of the XML/ OODB-processor itself will be taken. © Springer-Verlag Berlin Heidelberg 2005.