Dublin Core - Tagging the Web for better search and retrieval (1/3) - exploring XML
Dublin Core - Tagging the Web for better search and retrieval
The Web is a vast, but rather unstructured resource for information of any kind. While the lack of structure allowed for such a tremendous growth in the first place, it has recently turned more and more into a problem when trying to research specific topics and find relevant documents.
While search engines try to alleviate the problem and offer a simple interface to searching, the results can be sometimes unsatisfying. A crisis for search and retrieval has arrived because even all search engines taken together only cover a fraction of the Web, which is itself still growing by leaps and bounds every day. Correctly indexing the covered content is yet another challenge.
Meta data can be a solution to the latter problem, and XML is the grammar of choice for meta data. With DTDs or Schemas it is possible to define a document's structure, but not necessarily semantics, though. Giving elements human-readable names allows the human eye to more or less correctly interpret the purpose of a certain XML document, but the XML or HTML parser still has no idea.
What is called for is a standardized vocabulary for resource descriptions, in that not only a DTD or Schema is defined but also its elements and attributes have well-understood meanings. A group of SGML veterans is working on various aspects of the above-mentioned problems and defined a controlled vocabulary under the name of Dublin Core (DC).
Created: Nov 20, 2000
Revised: Nov 20, 2000