Promoting RDF (was RE: A certain difficulty)

For my part, the best way to get people to use RDF is to make sure as much
information (that people would value) as possible is available via RDF.  I'd
bet that for many applications, the tools to *use* the RDF information are
relatively straightforward given concrete examples to work from -- general
RDF tools must be much more difficult to design and implement.  Given useful
information, the tools should follow.

The other important tactic would be to highlight the use of RDF in other
applications.  For example, the entire Web syndication/weblogging movement
is based, in part, on the RSS method for describing site changes.  RSS, in
turn, was intended to be an RDF application (although I believe both have
diverged) and as such, RDF advocates can take a small share of credit for
the hundreds and thousands of interconnected Weblogs with (relatively)
seamless content syndication and aggregation.

In general, I'd personally to like see RDF used to link or embed additional
information in pages from database-driven Web sites.  Given a template-based
page generator, it should be a manageable effort to produce pages with both
human- and machine-usable content.  For example, the University of
Maryland's SHOE system provides a straightforward way to embed information
(in a near-XML way) in HTML pages for subsequent use by agents and other
services; RDF could provide a similar capability in a standards-based
fashion.  Think of it as "knowledge closed-captioning" and as one more way
to enable a machine-understandable/"actionable" Web.

Ken Meltsner
Knowledge Management, CA Services

Received on Friday, 25 February 2000 13:38:06 UTC