RE: ANNOUNCEMENT: Metalog v2.0b now available

Richard, there is admittedly a lack of complete technical documentation 
(this is in the todo's list..) regarding the current PNL.
The best public tech description that is available so far regarding the
PNL parsing rules can be found for example in 
http://www.w3.org/RDF/Metalog/ -> 
http://www.w3.org/RDF/Metalog/history.html ->
http://www.w3.org/RDF/Metalog/CIKM-050299.html (Towards the Semantic Web: Metalog)
(or the other previous versions, essentially the same for the PNL parsing).

That refers to the v1 incarnation, there are some minor differences in the v2
which are hinted at in the internal Metalog documentation (available as html files).
Anyway, yes, the plan is definitely to write more details on this.
In fact, the best thing would be probably to add similar approaches to the PNL page
http://www.w3.org/RDF/Metalog/docs/pnl.html (or some aggregate of it), so 
to better describe all the approaches that try to climb up the P axis using
PNL-like approaches (or other approaches). That is part of the "better documentation" 
issue (time allowing..). This would of course include MKR in view of its 
n-triple import, and RDF tabula rasa built-in.
I didn't do a formal study, but as far as drawing MKR's point in the 
diagram, it would be likely strictly more expressive than Metalog,
sitting lower in the P axis, so filling up another part of the P-space.
This relates to Danbri's question on relationships to CWM as well
(formal inter-relationships), CWM's point would be in a similar situation,
maybe inbetween the two (in a hierarchy), maybe elsewhere (again, didn't
do a formal check yet). Will try find some time to elaborate on this (or, 
someone step in ;)

Hear you soon,
-M





> -----Original Message-----
> From: Richard H. McCullough [mailto:rhm@cdepot.net]
> Sent: Wednesday, October 15, 2003 11:02 AM
> To: Massimo Marchiori
> Cc: RDF-Interest
> Subject: Re: ANNOUNCEMENT: Metalog v2.0b now available
> 
> 
> I'm very interested in PNL.
> Do I need to download Metalog to get the details of PNL?
> 
> FYI, my MKR language (http://rhm.cdepot.net/)
> is English-like (not as close to NL),
> unambiguous (context supplies the details) 
> and unrestricted (you can say anything).
> 
> Your PNL example
>     JOHN and MARY OWN a "red house". 
> 
> is expressed in MKR as
>     John, Mary do own od a red house done;
> 
> with the context including propositions like
>     John, Mary isa person;
>     John, Mary do marry done;
>     red isa color;
>     ...
> 
> Dick McCullough
> knowledge := man do identify od existent done;
> knowledge haspart proposition list;
> http://rhm.cdepot.net/
> ----- Original Message ----- 
> From: "Massimo Marchiori" <massimo@w3.org>
> To: "RDF-Interest" <www-rdf-interest@w3.org>
> Sent: Tuesday, October 14, 2003 5:45 PM
> Subject: ANNOUNCEMENT: Metalog v2.0b now available
> 
> 
> > 
> > A new release of the Metalog distribution (v2.0b) is now available.
> > 
> > Metalog home page at http://www.w3.org/RDF/Metalog/
> > Direct download page at http://www.w3.org/RDF/Metalog/download/
> > 
> > Enjoy,
> > -M
> 
>

Received on Wednesday, 15 October 2003 14:00:17 UTC