W3C home > Mailing lists > Public > public-owl-dev@w3.org > April to June 2009

Re: OWL2 questions

From: Thomas Loertsch <loertsch.thomas@guj.de>
Date: Tue, 12 May 2009 18:11:20 +0200
To: Alan Ruttenberg <alanruttenberg@gmail.com>
CC: public-owl-dev <public-owl-dev@w3.org>
Message-ID: <C62F6C48.1267%loertsch.thomas@guj.de>
i revised the ontology on http://purl.org/arecipe/ according to your advice
(i hope). a typical element now looks like the example below. so it's indeed
two elements. to me this looks rather appealing since it seperates the
somehow exaggerated datatyping constructs into a separate class while the
base class sticks to the customs and remains easily read- and query-able.
additionally i now have a nice hook for the comment without needing to reify
anything. there's still the non-standard datatype
<http://www.w3.org/TR/html/> because the regex will take a little longer.

again comments appreciated! thanks,
thomas


:instructions
  rdf:type           owl:DatatypeProperty ;
  rdfs:label         "instructions" ;
  rdfs:isDefinedBy   <http://microformats.org/wiki/hrecipe#instructions> ;
  rdfs:domain        :Recipe ;
  rdfs:range         xsd:string ;
  rdfs:comment
    "The method of the recipe."
  .

owl:Thing
  rdfs:subClassOf    [ owl:onProperty :instructions ;
                       rdf:type owl:Restriction ;
                       owl:allValuesFrom [
                         rdf:type rdfs:Datatype ;
                         owl:unionOf ( xsd:string
                                       <http://www.w3.org/TR/html/> )
                         ] ;
                       rdfs:comment
                         "The field MAY include valid HTML markup e.g.
paragraphs or a list of steps."
                       ] .






On 10.05.09 06:08, "Alan Ruttenberg" <alanruttenberg@gmail.com> wrote:

> On Tue, May 5, 2009 at 9:21 AM, Thomas Loertsch <loertsch.thomas@guj.de>
> wrote:
>> hello list,
>> 
>> i'm working on a mapping of microformats to rdf. i'm trying to capture all
>> the properties of every element of every microformat as well as the
>> restrictions on element values. the task is bigger than i had thought
>> (hoped) and especially the modelling of the value constraints is rather
>> tricky. i have no local OWL guru whom i could pester with detailed questions
>> so i'm turning to you, list.
>> the full draft of the mapping can be found at http://purl.org/amicroformat
> 
> This url doesn't resolve - you need to have a trailing "/"
> 
> http://purl.org/amicroformat/
> 
>> (a N3 version ready for human consumption as well as a machine generated
>> rdf/xml version) but that's a long read. i would be very grateful if someone
>> could comment on the following constructs. are they correct? do they express
>> what the rdfs:comments suggest? did i get the use of unionOf and
>> intersectionOf right? or should i have used oneOf? and what about the blank
>> nodes - do they "sit" right?
>> 
>> 
>> the method element for example i still quite simple. the original
>> microformat says "The field MAY include valid HTML markup e.g. paragraphs or
>> a list for steps in the method." so both xsd:strings and html are correct.
>> that's why i used unionOf
>> 
>> 
>> :method
>>  a  owl:DatatypeProperty ;
>>  rdfs:label  "method" ;
>>  rdfs:isDefinedBy <http://microformats.org/wiki/hrecipe#method> ;
>>  rdfs:domain :Recipe ;
>>  rdfs:range
>>    [ a owl:Restriction ;
>>      owl2:onDatatype
>>      [ owl:unionOf
>>         ( xsd:string
>>           <http://www.w3.org/TR/html/> )
>>        ] ;
>>      rdfs:comment "The field MAY include valid HTML markup e.g. paragraphs
>> or a list for steps in the method."
>>                       ] ;
>>  rdfs:comment
>>    "The method of the recipe."
> 
> Find below some translations of the above. Some notes:
> 
> - I give two versions as regards the second rdfs:comment as I wasn't
> sure whether you wanted the comment "The field ...." to be an
> annotation on :method or on the axiom asserting the range. I think the
> correct one is to have the comment on :method, as the subject of the
> comment is "The field", rather than the range axiom.
> 
> - One doesn't use owl:Restriction in this place - it is used in
> certain class expressions, rather than in asserting a range or domain.
> Hoever, it is the case that DataPropertyRange( DPE DR ) means the same
> thing as SubClassOf( owl:Thing DataAllValuesFrom( DPE DR ) ), and in
> that case one would use owl:Restriction.  I give an example of this
> formulation below (3)
> 
> - There is no owl2 namespace, just owl.
> 
> - I don't think <http://www.w3.org/TR/html/> denotes a datatype, as
> you would have it. Seems to me that it is a document. Be aware that
> syntactically this is OK, but that it may not follow common practice,
> and would take your ontology out of OWL DL as it is not one of the
> supported datatypes. Closer in spirit would be the mime type for html,
> but a) I am unaware of a standard URI for this, and b) There is the
> same problem about the datatype not being in the OWL 2 datatype map.
> I'm afraid that you might be best off having the range be string, and
> having a comment about the html.
> 
> If you can define a regular expression that matches html
> ;-) , you could define an html type by create a datatype restriction
> on xsd:string with a pattern facet. But in that case you might as well
> again write xsd:string as the range as the union of a superclass and
> one of it's subclasses is equivalent to the superclass.
> 
> - In each of the case I give the OWL 2 Functional-style syntax,
> followed by a turtle translation, for your reference.
> 
> 1) comment on :method
> 
> == Functional-style syntax ==
> 
> Declaration(DataProperty(:method))
> AnnotationAssertion(rdfs:isDefinedBy :method
> <http://microformats.org/wiki/hrecipe#method>)
> AnnotationAssertion(rdfs:comment :method "The method of the recipe.")
> AnnotationAssertion(rdfs:comment :method "The field MAY include valid
> HTML markup e.g. paragraphs or a list for steps in the method.")
> DataPropertyDomain(:method :recipe)
> DataPropertyRange(:method DataUnionOf(xsd:string <http://www.w3.org/TR/html/>)
> 
> == Turtle syntax ==
> 
> :method
>       rdf:type owl:DatatypeProperty ;
>       rdfs:comment "The method of the recipe." ,
>           "The field MAY include valid HTML markup e.g. paragraphs or
> a list for steps in the method." ;
>       rdfs:domain :recipe ;
>       rdfs:range
>               [ rdf:type rdfs:Datatype ;
>                 owl:unionOf (xsd:string <http://www.w3.org/TR/html/>)
>               ] ;
>       rdfs:isDefinedBy <http://microformats.org/wiki/hrecipe#method> .
> 
> 2) comment on the range axiom
> 
> == Functional-style syntax ==
> 
> Declaration(DataProperty(:method))
> AnnotationAssertion(rdfs:isDefinedBy :method
> <http://microformats.org/wiki/hrecipe#method>)
> AnnotationAssertion(rdfs:comment :method "The method of the recipe.")
> DataPropertyDomain(:method :recipe)
> DataPropertyRange(
>  Annotation(rdfs:comment "The field MAY include valid HTML markup e.g.
> paragraphs or a list for steps in the method.")
>  :method
>  DataUnionOf(xsd:string <http://www.w3.org/TR/html/>))
> 
> == Turtle syntax ==
> 
> :method
>       rdf:type owl:DatatypeProperty ;
>       rdfs:comment "The method of the recipe." ;
>       rdfs:domain :recipe ;
>       rdfs:range _:b1 ;
>       rdfs:isDefinedBy <http://microformats.org/wiki/hrecipe#method> .
> 
> _:b1  rdf:type rdfs:Datatype ;
>       owl:unionOf (xsd:string <http://www.w3.org/TR/html/>) .
> 
> []    rdf:type owl:axiom ;
>       rdfs:comment "The field MAY include valid HTML markup e.g.
> paragraphs or a list for steps in the method." ;
>       owl:subject :method .
>       owl:predicate rdfs:domain ;
>       owl:object _:b1 ;
> 
> 3) A form that would use owl:Restriction (comment on :method)
> 
> == Functional-style syntax ==
> 
> Declaration(DataProperty(:method))
> AnnotationAssertion(rdfs:isDefinedBy :method
> <http://microformats.org/wiki/hrecipe#method>)
> AnnotationAssertion(rdfs:comment :method "The method of the recipe.")
> AnnotationAssertion(rdfs:comment :method "The field MAY include valid
> HTML markup e.g. paragraphs or a list for steps in the method.")
> DataPropertyDomain(:method :recipe)
> SubClassOf(owl:Thing DataAllValuesFrom(:method DataUnionOf(xsd:string
> <http://www.w3.org/TR/html/>))
> 
> == Turtle syntax ==
> 
> owl:Thing
>       rdfs:subClassOf
>               [ rdf:type owl:Restriction ;
>                 owl:allValuesFrom
>                         [ rdf:type rdfs:Datatype ;
>                           owl:unionOf (xsd:string
> <http://www.w3.org/TR/html/>)
>                         ] ;
>                 owl:onProperty :method
>               ] .
> 
> :method
>       rdf:type owl:DatatypeProperty ;
>       rdfs:comment "The method of the recipe." ,
>          "The field MAY include valid HTML markup e.g. paragraphs or a
> list for steps in the method." ;
>       rdfs:domain :recipe ;
>       rdfs:isDefinedBy <http://microformats.org/wiki/hrecipe#method> .
> 
> I will look in to the next example and try to respond later. The
> appropriate references to study for the above are
> 
> http://www.w3.org/TR/2009/WD-owl2-syntax-20090421/
> http://www.w3.org/TR/2009/WD-owl2-mapping-to-rdf-20090421/
> 
> Note that OWL 2 is currently in last call. If you read these documents
> and have comments, please send them to public-owl-comments@w3.org by
> May 13, 2009.
> 
> -Alan
> 
> http://sciencecommons.org/about/whoweare/ruttenberg/
> 
> 
> 
>> the second example is harder to model: the hcard-conventions have to be
>> followed and certain html-elements are optional. i used a combination of
>> intersectionOf and unionOf (i wouldn't know how to model the difference
>> between "SHOULD" and "MAY" though).
>> 
>> 
>> :photo
>>  a owl:ObjectProperty;
>>  rdfs:label "photo" ;
>>  rdfs:isDefinedBy <http://microformats.org/wiki/hcard> ;
>>  rdfs:seeAlso <http://microformats.org/wiki/hrecipe#photo> ;
>>  rdfs:domain :Card ,
>>              :Recipe;
>>  rdfs:range
>>    [ a owl:Restriction ;
>>      owl2:onDatatype
>>      [ owl:intersectionOf
>>        [ owl:unionOf
>>          ( xsd:string
>>            <http://www.w3.org/TR/REC-html40/struct/objects.html#h-13.2>
>>            <http://www.w3.org/TR/REC-html40/struct/links.html#h-12.2>
>>            <http://www.w3.org/TR/REC-html40/struct/objects.html#h-13.3>
>>          ) ,
>>          mfs:ACard
>>        ] ;
>>        rdfs:comment
>>         "Typically used with an <img> tag. Use the 'src' attribute for URI
>> values. Use the 'data:' URI scheme for binary values.
>> <http://www.ietf.org/rfc/rfc2426.txt>" ,
>>         "The element SHOULD use an <img> element. The element MAY use any
>> other element that contains a URL, such as <a> or <object>, but it is not
>> recommended.  The contents of the element MUST follow the conventions
>> outlined in <a href='http://microformats.org/wiki/hCard'>hCard</a>."
>>      ]
>>    ] ;
>>  rdfs:comment
>>    "See section 3.1.4 of RFC 2426. <http://www.ietf.org/rfc/rfc2426.txt>" ,
>>    "Accompanying image. "
>>  .
>> 
>> 
>> any comments greatly appreciated!
>> thomas
>> 
>> 
>> 
>> .
>> Thomas Lörtsch
>> Business Development
>> G+J Exclusive&Living digital GmbH
>> Redaktion Online
>> ..
>> Stubbenhuk 5
>> 20459 Hamburg
>> ...
>> eMail: loertsch.thomas@guj.de
>> 
>> 
>> 
>> 
>> 



.
Thomas Lörtsch
Business Development
G+J Exclusive&Living digital GmbH
Redaktion Online
..
Stubbenhuk 5
20459 Hamburg
...
eMail: loertsch.thomas@guj.de
Received on Tuesday, 12 May 2009 16:12:20 GMT

This archive was generated by hypermail 2.3.1 : Wednesday, 27 March 2013 09:32:57 GMT