Re: facts about web ontology languages

I'll put myself up for being shot down and claim to be knowledgeable!

Frank Clar wrote:
> Hello,
> 
> right now I am working on a semantic web project at my University and
> therefore I need to know some facts about the following web ontology
> languages RDF/S, DAML+OIL and OWL. I could not find an answer to my
> questions neither in the specifications of the standards nor in any
> scientific article. So I thought that maybe someone from this mailing
> list can help me. I hope I have chosen the right one for those kind of
> questions. 
> 
> Here are my questions:
>  
> Are they completely decidable?

OWL Full is not decidable

OWL DL, OWL Lite, RDF, RDFS (without datatypes other than 
rdf:XMLLiteral) decidable.
RDFS + datatyping not known. (I believe it to be, I expect decidability 
to be proved next year).

See Herman ter Horst's excellent ISWC2004 paper for RDFS result.

> Is it right that only OWL full and RDF/s are not decidable, because they
> do not seperate between concepts and instances?
> 

No.

e.g. OWL DL without the complicated restriction on transitivity 
properties is not decidable, and RDFS is decidable, hence classes as 
instances is orthogonal to decidability.

> Do DAML+OIL and OWL DL support Description logics and for this reason as
> well first order predicate logic? 
>

No DAML+OIL and OWL DL support DLs for social/hsitorical reasons, some 
of which have technical underpinnings. Certainly the work on 
decidability and tractability of fairly expressive DLs is impressive, 
but other choices could have been made.

> And do they support first respectively second order predicate logic?
> Only DAML+OIL and OWL lite and DL support first order predicate logic,
> because they are based on Description logics?

DAML+OIL is given an OWL Full like reading in the DAML+OIL axiomatic 
semantics, and an OWL DL like reading in the DAML+OIL model theoretic 
semantics. The DAML+OIL and OWL DL model theoretic semantics have a 
particular mapping into first order predicate logic. OWL Full has a 
different mapping into first order predicate logic.

> 
> What different kinds of syntax do the above named languages support and
> is it possible to map all of them to UML? 
> RDF/XML, abstract syntax, n3, n-triples

Yes, to syntaxes.

Note the OWL abstract syntax can also be used for a subset of OWL Full 
that is larger than OWL DL.

> ...I have just found several
> articles, which describe methods to map RDF/S to UML. 
> 

Don't know.

> Is it possible to express a kind of class variable, which has one single
> value for a concept? 

Yes, in OWL Full, no in OWL DL.

<owl:FunctionalProperty rdf:ID="singleton">
   <rdfs:domain rdf:resource="&owl;Class/>
</owl:FunctionalProperty>


Note two equivalent but different classes may have different values for 
the singleton property.

> Is there a way to define one indirectly?  

??

> 
> Do DAML+OIL and OWL support reification and do they offer the possiblity
> to add additional information to the statements? I guess that
> reification has no meaning in OWL and DAML+OIL. 

Reification should be avoided, the semantics of it are not what you 
might expect:

http://www.w3.org/TR/2004/REC-rdf-mt-20040210/#Reif

e.g.
[[
In sum: the meaning of the reification is that a document exists 
containing a triple token which means whatever the first graph 
means.Note that this way of understanding the reification vocabulary 
does not interpret reification as a form of quotation. Rather, the 
reification describes the relationship between a token of a triple and 
the resources that triple refers to. The reification can be read 
intuitively as saying "'this piece of RDF talks about these things" 
rather than "this piece of RDF has this form".
]]
and
[[
Since an assertion of a reification of a triple does not implicitly 
assert the triple itself, this means that there are no entailment 
relationships which hold between a triple and a reification of it. Thus 
the reification vocabulary has no effective semantic constraints on it, 
other than those that apply to an rdf-interpretation.
]]

Reification is particularly problematic in OWL DL, where you have to 
decide whether to try rdf:object as an owl:DatatypeProperty or an 
owl:ObjectProperty.

> 
> Could you declare a default value for a literal, if no value was
> assigned to this property?
> 

No. None of the languages provide any support for defaults.

> Thanks in advance!
> Best regards
> 
> Frank Clar
> 

Jeremy

Received on Tuesday, 14 December 2004 15:02:42 UTC