Improving Search Quality of an XML Context-Driven Search Engine by Relabling of Canonical Trees

Abstract

XCD Search-An XML Context-driven Search Engine answers both Keyword-based and Context-driven queries using stack-based sort merge algorithm. It performs well with all criteria of queries against XML trees, except queries submitted against a document, whose XML tree contains a parent node and child interior node, both having same Taxonomic Label and both have child/children data node(s) and/or attribute(s). In this paper we propose An Improved XML Context-driven Search Engine. It uses all the techniques used in XCD Search, in addition to new techniques that handle the type of XML trees mentioned above, which XCD Search does not handle well. We evaluated this system experimentally and compared with original version of XCD Search. The results showed remarkable improvement.