Re: OWL-DL and linked data

2008/6/10 Bernard Vatant <bernard.vatant@mondeca.com>:
>> In these previous discussions has it been mentioned that these
>> vocabularies seem to be excluding definitional resources?  Is linking
>> data supposed to include other representations?
>
> Clearly enough, the linked data space as it stands restricts itself to RDF
> data. Other representations seem out of scope, but are welcome to be
> migrated or otherwise wrapped in RDF..
>>
>> I mean other languages like Topic Maps or English should be part of idea
>> of linked
>> data
>
> Topic maps I can understand. English? What do you mean exactly by that.

Things can be defined by non-RDF resources and English is one of these
(or more correctly however it's delivered through HTML, PDF, etc).  If
it was a format that could be easily converted to RDF maybe that
should happen?  Properties, classes etc are documented in ontologies
with annotations this is just another, more indepth, extensible type.

It would be similar to:
http://neurocommons.org/page/URI_documentation_protocol

However, it puts the link(s) in the RDF so the client doesn't have to
parse the HTML or do other HTTP requests (it eliminates step 5).
Maybe an isDefinedBy that allows content negotiation (doesn't restrict
RDF)?

Received on Wednesday, 11 June 2008 03:20:44 UTC