ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] Last Call: OWL 2 and rdf:text primitive datatype

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Joel Bender <jjb5@xxxxxxxxxxx>
Date: Mon, 27 Apr 2009 14:21:00 -0400
Message-id: <49F5F78C.9040109@xxxxxxxxxxx>
John,    (01)


I'm not sure I can eliminate some of your bumfuzzledness, but here is my 
perspective.    (02)

> Sometimes that is the intension, I hope that is not the case here.    (03)

As much of a fan of conspiratorial obfuscation as I am, I really don't 
think this is the case.    (04)

> Specifically, I notice some things about the overview document.    (05)

A common expression in my office: You can't get there from here, you 
have to go down the road a bit.    (06)

> 1. The figure 1 image is said to show the language. It doesn't
>     look like any language structure I've seen.    (07)

It's not the language, it's the relationship between the concepts being 
presented and a variety of optional and required serializations of those 
concepts.    (08)

> Normally, I would expect to see some initial assumptions or
> operating characteristics and perhaps a BNF or other more formal
> representation.    (09)

To get that, look at section 2.2, which refers to direct semantics, 
which leads you to: <http://www.w3.org/TR/owl2-semantics/>    (010)

> I would also expect to see a few examples  to show how those
> language features are used.    (011)

For that, check out the primer <http://www.w3.org/TR/owl2-primer/> and 
the <http://www.w3.org/TR/owl2-new-features/> documents.    (012)

> 2. The ellipse shows both "ontology structure" and "RDF graph" as
> part of the language. Can that be right? Is the maping also part
> of the language?    (013)

Yes and no.  The mapping is more of a conceptual level relationship, 
rather than a syntactical one.  There are some concepts in OWL that fit 
easily with RDF and many more that don't.  I think of OWL as more like 
RDF+, or more like RDFS+, where the RDF serialization of some of these 
OWL constructs hasn't been completely ironed out yet.    (014)

> Parsing is shown to be possible for importing from a number of
> other document types. Again, are these part of the language?    (015)

No, but people that have backgrounds in other languages at least have 
something that says "if you understand something from this other world 
you will not be completely lost, may of the same concepts you love and 
cherish are here with other labels".    (016)

The truly nasty twist is that there are so many assumptions people bring 
along (as outlined in the extensional and intensional thread or this 
forum) don't fit, are designed to *not* fit, will *never* fit, and only 
after people have committed themselves to diving into OWL do they learn 
this.    (017)

> Are the parsers part of the language or are they external tools?    (018)

No more than C++ compilers are part of the C++ language.    (019)

> Direct (OWL 2?) semantics and RDF-Based Semantics are both shown
> as part of the diagram. I've worked on systems that attempted to
> include dual semantics and they were very difficult to keep
> synchronized. Am I misreading this?    (020)

This is more like translation between C and Pascal data structures. 
There are going to be lots of things that can go across the divide 
easily, some things can't.  For example, does the fact that Pascal have 
array bounds checking built in and that C does not factor into a 
language translation between the two for some specific structure definition?    (021)

See that cute little statement in small text above a dotted line?    (022)

     "correspondence theorem (for DL subset)"    (023)

It reminds me of the health inspector shouting in the Monty Python 
episode "I think it's be more appropriate if the box bore a great red 
label: "WARNING: LARK'S VOMIT!!!".    (024)

Here there be dragons.    (025)


Joel    (026)

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (027)

<Prev in Thread] Current Thread [Next in Thread>