Re: ITSRDf definitions

Hi Jirka,
[copying to list]

Thanks for this advice, and to Felix for similar input at the last call 
on the 4th March.

I'll then dump the generated xsd schema and do dedicate schema. In fact 
most of the types are URI or strings which we can declare directly, and 
just define new types for the enumerations, in a way that maps directly 
to the ITS types.

I presume we should design a single on to be used in both the NIF and 
the PROV-O mapping?

How does that sound?

Note that this also remove any xsd dependencies  with the XLIFF mapping.

cheers,
Dave

On 24/02/2013 21:43, Jirka Kosek wrote:
> On 24.2.2013 2:20, Dave Lewis wrote:
>
>> So here's my thinking on the its-rdf ontology, I'd welcome any comments
>> as I've not had a lot of time to spend on this:
>>
>> *<http:*//example.com/exampledoc.html#offset_0_29>
>>      itsrdf:translate
>> "yes"^^*<http:*//www.w3.org/TR/its-2.0/its.xsd#yesOrNo>.
>>
>> I wasn't sure where this was grounded, since we don't
>> have*http:*//www.w3.org/TR/its-2.0/its.xsd
>>   Further if we did it would it not make sense to generate it from the
>> current
>>   relaxNG data types (annex D, part 4 of spec) using trang or equivalent.
>>
>> The attached xsd is the corresponding output of trang. You can see from
>> this that the
>> corresponding type name would not be "yesOrNo" but "its-translate.type"
>> (as trang takes the
>> XML schema name directly from the RELAX NG schema name.
> I personally don't think that it makes sense to tie together current
> schemas we have with ontology. RELAX NG schema we have was built mainly
> for validation and as a building block for another vocabularies hosting
> ITS 2.0. If for RDF definition of types is necessary in the form of XML
> Schema then I think it's better to create such schema manually. I don't
> think that type like itstype:its-translate.type improves readability of
> the schema.
>
>      Jirka
>

Received on Wednesday, 6 March 2013 11:06:49 UTC