W3C home > Mailing lists > Public > public-lod@w3.org > April 2010

Re: Using predicates which have no ontology?

From: Niklas Lindström <lindstream@gmail.com>
Date: Tue, 6 Apr 2010 16:59:35 +0200
Message-ID: <q2rcf8107641004060759r2a79eca4n458affb11b7f950e@mail.gmail.com>
To: Michael Hausenblas <michael.hausenblas@deri.org>
Cc: nathan@webr3.org, Danny Ayers <danny.ayers@gmail.com>, Phil Archer <phil@philarcher.org>, Linked Data community <public-lod@w3.org>
Michael,

2010/4/6 Michael Hausenblas <michael.hausenblas@deri.org>:
>
>> If not, would you consider updating your interim solution to describe
>> URI:s under [1]? I mean, since [2] currently uses the real IANA URI:s
>> (i.e. the "unsanctioned" ones) and those, as Danny cautioned, could
>> end up e.g. being resolved to documents, breaking semantics (as well
>> as not being discoverable).
>
> I'm not totally sure if I understand but I guess the answer would be yes ;)

Nice. I mean to the effect of minting something like
<http://purl.org/NET/atom-link-rel#self> etc. for each of them. :)


> It's interesting that you've modelled the relation-type as RDF properties in
> [4] whereas I turned them (in [1]) into instances of the class
> 'awol:RelationType' from the AtomOwl vocabulary.
>
> Any thoughts?

Ah, yes. I just made a minimalistic set of statements about them, with
only RDF semantics. Via RDFS and OWL your statements entail mine,
since awol:RelationType is a rdfs:subClassOf owl:ObjectProperty (and,
via that, of rdf:Property).

If IANA wants to use RDF to define the link relations (which I assume
all of us here think they should), the question remains how rich
semantics they're willing to add to the definitions.)

Best regards,
Niklas



> Cheers,
>      Michael
>
> --
> Dr. Michael Hausenblas
> LiDRC - Linked Data Research Centre
> DERI - Digital Enterprise Research Institute
> NUIG - National University of Ireland, Galway
> Ireland, Europe
> Tel. +353 91 495730
> http://linkeddata.deri.ie/
> http://sw-app.org/about.html
>
>
>
>> From: Niklas Lindström <lindstream@gmail.com>
>> Date: Sat, 3 Apr 2010 22:46:29 +0200
>> To: Michael Hausenblas <michael.hausenblas@deri.org>
>> Cc: <nathan@webr3.org>, Danny Ayers <danny.ayers@gmail.com>, Phil Archer
>> <phil@philarcher.org>, Linked Data community <public-lod@w3.org>
>> Subject: Re: Using predicates which have no ontology?
>>
>> Hi Michael,
>>
>> that's great! If [2] were to be updated with that [1] (i.e. officially
>> containing RDFa about these URI:s), and would be 303:d to from [3]
>> (along with anything under that URL), this would be all we need. I
>> know it hasn't happened for years, but sometimes a nudge at just the
>> right time may be all it takes..
>>
>> If not, would you consider updating your interim solution to describe
>> URI:s under [1]? I mean, since [2] currently uses the real IANA URI:s
>> (i.e. the "unsanctioned" ones) and those, as Danny cautioned, could
>> end up e.g. being resolved to documents, breaking semantics (as well
>> as not being discoverable).
>>
>> I did a manual (well, vim-macro:ed) conversion of [3] into RDF/XML,
>> but had to leave to eat easter eggs at my sister's and entertain her
>> kids. :) It's located at [4] now, and quite similar to the data in
>> [1]. Note that I do consider [1] much more interesting.
>>
>> (That said, if anyone would like me to make e.g. an XSLT for turning
>> [4] into something like [1], just say the word.)
>>
>> Best regards and happy easter!
>> Niklas
>>
>> [1]: <http://purl.org/NET/atom-link-rel>
>> [2]: <http://www.iana.org/assignments/link-relations/link-relations.xhtml>
>> [3]: <http://www.iana.org/assignments/relation/>
>> [4]: <http://bitbucket.org/niklasl/tripleheap/src/tip/iana-link-relations.rdf>
>>
>>
>> On Sat, Apr 3, 2010 at 8:38 PM, Michael Hausenblas
>> <michael.hausenblas@deri.org> wrote:
>>>
>>> Nathan, Phil, All,
>>>
>>>> and quote:
>>>> "If the relation-type is a relative URI, its base URI MUST be
>>>>    considered to be "http://www.iana.org/assignments/relation/"
>>>> http://tools.ietf.org/id/draft-nottingham-http-link-header-03.txt
>>>>
>>>> obviously all the links defined by:
>>>> http://www.iana.org/assignments/link-relations/link-relations.xhtml
>>>> (from the atom rfc)
>>>>
>>>>>> such as edit, self, related etc - with additional consideration to the
>>>>>> thought that these will end up in rdf via RDFa/grddl etc v soon if not
>>>>>> already.
>>>>>>
>>>>>> Any guidance?
>>>
>>> Yes. Use [1] ...
>>>
>>> My motto is: acting rather than talking. So, I took [2] as a starting point
>>> - which is already in nice XHTML format - and manually added some RDFa.
>>> After an hour I ended up with [1] (though, to be fair, two Wii games with
>>> the kids and consuming some Easter eggs also took place in that hour).
>>>
>>> So, [1] is really a sort of an interim solution (though, in the distributed
>>> data world I do expect much more of such fixes) and I encourage Phil, who is
>>> an editor of [2] to use the template from [1] at the 'official' location.
>>>
>>> Happy Easter! (and back to Wii games, for now ;)
>>>
>>> Cheers,
>>>      Michael
>>>
>>> [1] http://purl.org/NET/atom-link-rel
>>> [2] http://www.iana.org/assignments/link-relations/link-relations.xhtml
>>>
>>> --
>>> Dr. Michael Hausenblas
>>> LiDRC - Linked Data Research Centre
>>> DERI - Digital Enterprise Research Institute
>>> NUIG - National University of Ireland, Galway
>>> Ireland, Europe
>>> Tel. +353 91 495730
>>> http://linkeddata.deri.ie/
>>> http://sw-app.org/about.html
>>>
>>>
>>>
>>>> From: Nathan <nathan@webr3.org>
>>>> Organization: webr3
>>>> Reply-To: <nathan@webr3.org>
>>>> Date: Sat, 03 Apr 2010 00:14:16 +0100
>>>> To: Danny Ayers <danny.ayers@gmail.com>
>>>> Cc: Linked Data community <public-lod@w3.org>
>>>> Subject: Re: Using predicates which have no ontology?
>>>> Resent-From: Linked Data community <public-lod@w3.org>
>>>> Resent-Date: Fri, 02 Apr 2010 23:14:54 +0000
>>>>
>>>> Danny Ayers wrote:
>>>>> On 3 April 2010 00:53, Nathan <nathan@webr3.org> wrote:
>>>>>> Hi All,
>>>>>>
>>>>>> Any guidance on using predicates in linked data / rdf which do not come
>>>>>> from rdfs/owl. Specifically I'm considering the range of:
>>>>>>  http://www.iana.org/assignments/relation/*
>>>>>
>>>>> Can't find a URL that resolves there
>>>>
>>>> snap; but that's what rel="edit" and so forth resolves to.
>>>>
>>>> see example:
>>>> http://www.w3.org/2001/tag/doc/selfDescribingDocuments.html#ATOMSection
>>>>
>>>> and quote:
>>>> "If the relation-type is a relative URI, its base URI MUST be
>>>>    considered to be "http://www.iana.org/assignments/relation/"
>>>> http://tools.ietf.org/id/draft-nottingham-http-link-header-03.txt
>>>>
>>>> obviously all the links defined by:
>>>> http://www.iana.org/assignments/link-relations/link-relations.xhtml
>>>> (from the atom rfc)
>>>>
>>>>>> such as edit, self, related etc - with additional consideration to the
>>>>>> thought that these will end up in rdf via RDFa/grddl etc v soon if not
>>>>>> already.
>>>>>>
>>>>>> Any guidance?
>>>>>
>>>>> By using something as a predicate you are making statements about it.
>>>>> But...
>>>>>
>>>>> If you can find IANA terms like this, please use them - though beware
>>>>> the page isn't the concept. You might have to map them over to your
>>>>> own namespace, PURL URIs preferred.
>>>>
>>>> Would it make sense to knock up an ontology for all the standard
>>>> link-relations and sameAs them through to the iana uri's?
>>>>
>>>> Best, Nathan
>>>>
>>>
>>>
>>>
>
>
Received on Tuesday, 6 April 2010 15:00:28 UTC

This archive was generated by hypermail 2.3.1 : Sunday, 31 March 2013 14:24:26 UTC