Introducing Information Toolkit, a new column for The Owl …
Metadata Matters
The term "metadata" gets tossed about a great deal these days, but few seem to understand what it means. To add to the confusion, sometimes the term "metadata tags" is shortened to "meta tags". Regardless, metadata is simply data about data. Metadata is a familiar topic to libraries, since cataloging information is metadata; however, in the electronic environment, the term increasingly refers to data that aids in the identification, description, and location of networked electronic resources. Metadata used in web pages describes the content of the web page, but does not appear on the web page itself. In this case, it functions somewhat as cataloging in publication data appears within a book. The use of metadata tags is gradually increasing, but many pages still do not have them.
Metadata tags typically are coded in the header of an HTML document, and often appear like these examples from the Kornhauser home page:
You can view metadata tags by opening the source code of a web page. In Netscape, go to the "view" menu and choose "page source." In Explorer, go to "view" then to "source."
As these examples illustrate, metadata tags contain information on the content and description of web pages. Many search engines use this information when indexing, retrieving, and ranking documents. Pages with metadata tags are more likely to be retrieved and have a higher ranking than pages without metadata tags. Metadata tags also add descriptive content to pages with little text, such as a page with lots of graphics. Pages with JavaScript or frames are often not indexed by search engines; metadata tags can improve indexing and retrieval in these cases.
Most pages that have metadata use it in the simplified "keyword" "content" form above. Some pages include more information so that something akin to a cataloging record in encoded into the page. There are movements afoot to create more descriptive, standardized metadata formats. The most notable is the Dublin Core Initiative, developed by an international collaboration and hosted by OCLC. The Dublin Core Initiative attempts to create an organizational infrastructure for the Internet. It goes far beyond "keyword" and "content" to include "creator" "date" "title" "publisher" and other descriptive terms. To explore more about the Core, see
http://purl.oclc.org/dc/index.htm.Interestingly, all search engines, including Lycos™ and Google™ do not support metadata. Don’t be fooled by these few exceptions; metadata is becoming an increasingly prominent and important topic within information science. It is a means to organize and control web pages, as well as an aid to document retrieval and ranking. If your interest is piqued, check out
http://www.searchenginewatch.com for more information including some lawsuits over metadata content, from Playboy, no less! Clearly, metadata matters.