An XML search engine for the World-wide Web

Abstract

The Internet contains vast amounts of information and yet users find it difficult to find documents that match their queries using current HTML search engines. HTML is the problem because it does not allow authors to include extra information in the document that describes the actual content. XML (eXtensible Mark-up Language) evolved in order to overcome these shortcomings of HTML. It allows authors to mark-up their document content using content-specific tags. This report documents a project that consists of the creation of an XML-based search engine. This search engine finds XML and DTD resources on the Internet and indexes their contents. The results are stored in a central index. The engine primarily uses the extra structure and information implied by the mark-up tags in XML documents in order to rank the index entries when a query is received. The structure of the document provides the search engine with the notion of contexts: the search engine can now detect whether there is a relation between words in the document and the extent of that relationship. This gives the engine a decisive advantage over the current collection of HTML search engines. This enables easier and more successful searching for the end-user