Re: Cambridge Communique

Hi Ken,

----- Original Message -----
From: Ken Laskey <kenneth.j.laskey@saic.com>
| Here is another opinion of the relationship.  Imagine an XML Schema
| that defines the "hard" metadata for a book, e.g. author, title,
| ISBN.  These are immutable properties of the book.  Now, the New York
| Times does a review of the book which creates "assertions" which are
| not hard facts but still related information.  The connection of
| these assertions to the book would be through RDF.  RDF might also be
| used by various booksellers to make assertions about the price they
| are individually charging or possibly special offers about the book.
|....

You touch on the thing I like best in RDF: anyone can make statements
about a resource.  There is no need to edit a central document.

Naturally, this means RDF statements can conflict.  As you say, some
data is "harder" than others.  However, I don't see anything stopping us
from encoding "hard" data in RDF along with the "soft". (We are saying
that a NYT book review is "soft" - I hope no NYT people read this.)

There would advantages in encoding both in RDF.  If desired, we could
make second order statements indicating which statements are considered
"hard" and which are "soft".  Or I might think there is a spelling
mistake in the Author's name (supposedly "hard" data) and I could write
a second order statement to this effect.

--
Arnold deVos
Langdale Consultants
adv@langdale.com.au

Received on Thursday, 11 November 1999 15:08:05 UTC