Assertion by reference [was: Comments on Annotated DAML...]

Grit Denker wrote:
[...]
> Questions/comments on http://www.daml.org/2000/10/daml-ont.daml :
[...]
> o Assertion by reference, importing
> 
>   The comment for the "imports" property reads
> 
>     "X asserts the content of Y by reference"
>     "if imports(X, Y) and you believe X and Y says something, then you
>      should believe it"
> 
>   What does "asserts" and "believe" mean in this context?

Good question ;-)

By way of example, consider a meeting invitation that
comes to you by way of email. Or rather: to your automated
secretary agent. You have it configured to "believe,"
i.e. to store into its database or otherwise act on,
messages that bear the address fred@example.com
in thier From: field.

So if fred@example.com sends you an invitation 
that asserts
	"A meeting propose is scheduled
	for for 4:30pmEST on Thu, 12 Oct 2000.
	You are invited."

i.e.
	<Meeting>
	  <scheduledFor>2000-10-12T1830Z</scheduledFor>
	  <inviteeEmail resource="mailto:denker@csl.sri.com"/>
	</Meeting>

your secretary-bot will dutifully
color that box yellow (meaning: you've been invited
to something at this time) on your calendar page.

On the other hand, your secretary-bot will not
do that for just any old message that contains
such an assertion. It won't "believe" messages
from sources that it hasn't been instructed/configured
to believe. It will just ignore other messages.


So that's what it means, in practice, to "assert"
and to "believe" something. Now... for assertion
by reference...

Fred's message might just say:

	Fred, please join the Frotz group;
	Their schedule is at
		http://frotz.example.org/schedule

Clearly Fred means to "endorse" that schedule,
i.e. to vouch for its veracity, i.e. to
assert its contents by reference, in such
a way that your secretary-bot should "believe" it
and color yellow any boxes that correspond
to scheduled meeting times of the Frotz group.

So... I might transcribe his message ala:

	<Invitation about="">
	  <imports resource="http://frotz.example.org/schedule"/>
	</Invitation>

	<Group>
	  <name>Frotz</name>
	  <schedule resource="http://frotz.example.org/schedule"/>
	  <inviteeEmail resource="mailto:denker@csl.sri.com"/>
	</Group>

Does that make sense? I'm not sure I've really motivated
it clearly. I think I'll have more to say once I've
built some tools that make use of this sort of information.


>   Is there any semantics that explains this?

Well... the idiom is inspired by the SHOE use-ontology idiom:

4.2 Declaring Ontology Usage
http://www.cs.umd.edu/projects/plus/SHOE/spec.html#DeclaringOntologyUsage

when I started working on converting SHOE to RDF,
Jeff Heflin explained:

"SHOE use-ontology statements are similar to the way RDF use
namespaces to identify schemas. I say similar and not equivalent because
a use-ontology says that I specifically agree only to the semantics
implied by this ontology and those ontologies which it extends."
Re: Converting SHOE to RDF: about 2/3 done; some gotchas
From: Jeff Heflin (heflin@cs.umd.edu)
Date: Thu, May 11 2000 
http://lists.w3.org/Archives/Public/www-rdf-interest/2000May/0093.html

and after a bit of back-and-forth, I agreed that what
he's talking about is something TimBL had written about...

	Assertion of another document
	http://www.w3.org/DesignIssues/Toolbox#Assertion

and since then, I have been working on integrating
this notion into a model for sending messages containing
assertions (and questions and requests, I hope) around
in the network:

	An RDF Model for GET/PUT and Document Management
	http://www.w3.org/2000/07/document-maintenance/
	Thu, 24 Aug 2000 14:23:14 GMT

which is, in turn, based on some other scribblings:

	Web Architecture: Protocols for State Distribution
	http://www.w3.org/Architecture/state
	Fri, 17 Sep 1999 21:28:26 GMT

I have been using larch to try to formalize it; I got
a fair amount of the essentials of HTTP formalized, but
I haven't been back to work on the rest for a while:

http://www.w3.org/XML/9711theory/HTTP
http://www.w3.org/XML/9711theory/HTTP.html
http://www.w3.org/XML/9711theory/HTTP.lsl

Oh... I did a rough presentation of this specification
of HTTP using larch at WWW9:

Specifying Web Architecture with Larch 
Dan Connolly
9th International World Wide Web Conference
Amsterdam
May 2000 
http://www.w3.org/2000/Talks/www9-larch/all.htm

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Wednesday, 11 October 2000 03:06:49 UTC