[rdfweb-dev] ANN: Pet Profiles

[Danny Ayers]

Benjamin Nowack wrote:

>
>I'm sure we will see a new sciam article being published soon:
>
>   "The Semantic Web is an extension of the current web in which 
>   information is  given well-defined meaning, better enabling 
>   computers, people and pets to work in cooperation."
>    -- Tim Berners-Lee, James Hendler, Ora Lassila,
>       The Semantic Web, Phase IIb, Scientific American, May 2004
>  
>
Woof!

>
>  
>
>>[...]
>>Incidentally, it occurred to me that if the XSLT was applied outside of 
>>a browser, the resulting XHTML should be roundtrippable back to RDF/XML. 
>>But CSS, like a lot of folks, doesn't seem to like qnames in attributes, 
>>e.g. <div id="pet:Pet">  - an issues for RDF-in-XHTML, odd browser 
>>behaviour or am I missing something?
>>    
>>
>hmm, I don't know enough about xhtml, but isn't this id stuff there for
>locally unique names, i.e. that you can't use prefixes and double 
>colons? dunno..
>  
>
Woof! Sorry, try again - yep, good point. But if you can keep :'s but 
escape them (as Masahide pointed out) then it should make life easier.

>  
>
>><owl:Ontology rdf:about="http://xmlns.com/foaf/0.1/">
>>  <dc:title>Hedwig Pet Ontology</dc:title>
>>    
>>
>when I had a look yesterday evening, I think it was 'rdf:about=""', 
>which was the proper uri, I guess.
>  
>
Oop. I'd better check again...

>  
>
>><owl:Class rdf:about="Pet">
>>  <rdfs:label>Pet</rdfs:label>
>>  <rdfs:seeAlso rdf:resource="http://xmlns.com/wordnet/1.6/Pet"/>
>>    
>>
>I'm not sure, but I think that doing a seeAlso does not map your
>Pet to wn:Pet. It would just encourage a tool to read the Wordnet
>definition but wouldn't make it know that/how the two terms are
>related. I think you could use a subClassOf or (if you dare) an 
>equivalentClass construct to relate your Pet to Wordnet's Pet term.
>  
>
Yep, I wondered about this - I was terrified of equivalentClass, and 
even subClassOf seemed a little strong (although I'm now thinking that 
might be best). I messed around with the SKOS thesaurus stuff for a bit 
too - couldn't come up with anything there that seemed to fit either.

>><owl:Class rdf:about="Pet">
>>[...]
>>  <rdfs:label>Pet</rdfs:label>
>>[...]
>></owl:Class>
>><owl:Class rdf:about="PetOrPerson"> 
>>  <rdfs:label>Pet or Person</rdfs:label>
>>  <rdfs:comment>A pet or a person</rdfs:comment>
>>  <owl:unionOf rdf:parseType="Collection">
>>    <owl:Class rdf:about="http://xmlns.com/foaf/0.1/Person" />
>>  	<owl:Class rdf:about="#Pet" />
>>  </owl:unionOf>
>></owl:Class>
>>    
>>
>you are talking about (and describing) two Pet classes here,
>the rdf:about="Pet" expands to "http://purl.org/stuff/hedwig/Pet",
>the rdf:about="#Pet" to "http://purl.org/stuff/hedwig/#Pet".
>you could use an rdf:ID for the initial class definition and
>then use "#Pet" with other axioms. (+ #Dog, #Cat, #Human)
>  
>
Grrr.

>there are some minor namespace prefix typos (owl:subClassOf,
>owl:range)
>  
>
Thanks.

>what about making Pet a foaf:Agent. that would be a nice
>mapping and we could then reuse foaf:homepage, foaf:gender
>etc (unfortunately not foaf:knows for FOAP apps). 
>
Good idea.

>your 
>sample profile uses pet:name and pet:gender, which I 
>couldn't find in the spec..
>  
>
Ah, that's just an artifact of the (extreme just-in-time-or-a-bit-later 
dynamic make-it-up-as-you-go-along) methology I employed...

>and another funny "reasoning" thing resulting from
>"Human" being defined as a subClassOf "Pet":
>  
>
>><owl:Class rdf:ID="Pet"> (adjusted)
>>  <rdfs:comment>
>>     The class of animals kept for pleasure rather
>>     than utility
>>  </rdfs:comment>
>></owl:Class>
>>[...]
>><owl:Class rdf:about="Human">
>>  <rdfs:label>Human</rdfs:label>
>>  <rdfs:comment>
>>     The class of animals designated Homo sapiens
>>     (domestic human).
>>  </rdfs:comment>
>>  <rdfs:subClassOf rdf:resource="#Pet"/> (adjusted)
>></owl:Class>
>>    
>>
>
>What is the class extension of these
>    "Homo sapiens (domestic humans;), kept for
>    pleasure rather than utility"?
>
>let me see: strippers, callboys/-girls, children..
>  
>
Heh, I hadn't spotted that. "animals kept for pleasure" came out of a 
dictionary, but that had me a little worried too ;-)

>
>fun stuff. and the xsl/css is truly inspiring..
>  
>
Check out Masahide's iCal stuff (I can't find a link - anyone?). How it 
*should* be done ;-)

>benji
>   a pet:Dog ;)
>
>  
>
;-)

Thanks again.
Cheers,
Danny.