A method, computer readable medium and information processing system for projecting a representation of a document in the Extensible Markup Language (XML) format. The method dynamically creates a tree representation of an XML document based on a provided set of XPath expressions through approximate matching techniques. The projection created by the method supports all XPath axes including backward axes such as parent and ancestor. The execution of the set of XPath queries on the projected document is guaranteed to be the same as that of executing the XPath queries on the original document. The projected document typically occupies much less space than the original document.