Re: Wikidata export in RDF

Hi Denny,
I understand your point and in the work we did with OPM we didn't use blank
nodes either.
Your approach is perfectly valid as well.

Just an advice: if you qualify properties, always try to provide shortcuts
with unqualified
properties. This makes the qualification optional and makes querying the
model easier.

As a follow up of Jun's email, here is the latest draft of the DC to PROV
mapping note:
http://dvcs.w3.org/hg/prov/raw-file/1d34bb12f1a7/dc-note/Overview.html

Best,
Daniel

2012/8/9 Jun Zhao <jun.zhao@zoo.ox.ac.uk>

> FWIW, the WG is also working on a DC to PROV mapping note [1]. This is
> still a working draft. We would appreciate any further feedback from the DC
> community before we publish it.
>
> [1] https://dvcs.w3.org/hg/prov/**raw-file/ff940ee82d3d/dc-note/**
> Overview.html<https://dvcs.w3.org/hg/prov/raw-file/ff940ee82d3d/dc-note/Overview.html>
>
> Cheers,
>
> Jun
>
>
> On 08/08/2012 15:33, Antoine Zimmermann wrote:
>
>> DC does not intend to be compatible with OWL DL and its properties are
>> made to be flexible enough to be used with literals or URIs. However, I
>> like to consider DC properties as AnnotationProperties, and it seems
>> they are used as such by many people. Even in the OWL specifications, DC
>> properties are shown as typical examples of AnnotationProperties.
>> ProtĂŠgĂŠ 4 includes declaration of AnnotationProperties for some DC terms
>> by default. There are even two modified DC ontologies available where
>> all properties are declared as AnnotationProperties [1,2], which can be
>> used to "safely" import all the DC terms into a strictly OWL DL
>> ontology. These modified ontologies are compatible with OWL (1/2) DL.
>>
>> Nevertheless, in this particular case, I'd say it makes sense to use an
>> ObjectProperty instead, as it is core to the knowledge model of WikiData
>> and a source is always identified by a URI. AnnotationProperties should
>> be left for the things that are auxiliary to the domain being modelled.
>>
>>
>> [1] DC terms in OWL DL.
>> http://protege.stanford.edu/**plugins/owl/dc/terms.owl<http://protege.stanford.edu/plugins/owl/dc/terms.owl>
>> [2] DC terms in OWL 2 DL. http://purl.org/NET/dc_owl2dl/**terms<http://purl.org/NET/dc_owl2dl/terms>
>>
>>
>> Best,
>> AZ
>>
>> Le 08/08/2012 15:55, Markus KrĂśtzsch a ĂŠcrit :
>>
>>> On 08/08/12 14:28, Bob Ferris wrote:
>>>
>>>> Hi all,
>>>>
>>>> btw how about dct:source [1] or dct:publisher [2] as source (is derived
>>>> from) property?
>>>>
>>>
>>> Correct me if I am wrong, but I think prov has better OWL compatibility
>>> by declaring the type of its property (object vs. data). DC says that
>>> its properties are "intended to be used with non-literal values" but I
>>> am not sure if this is supposed to be a declaration as an object
>>> property or not.
>>>
>>> Cheers,
>>>
>>> Markus
>>>
>>>
>>>> Cheers,
>>>>
>>>>
>>>> Bo
>>>>
>>>>
>>>> [1] http://purl.org/dc/terms/**source <http://purl.org/dc/terms/source>
>>>> [2] http://purl.org/dc/terms/**publisher<http://purl.org/dc/terms/publisher>
>>>>
>>>> On 08/08/2012 02:33 PM, Denny Vrandečić wrote:
>>>>
>>>>> Hi Jun,
>>>>>
>>>>> almost perfect! prov:hadPrimarySource is a tiny bit too strong (it
>>>>> requires that source is from first-hand direct experience etc.),
>>>>> whereas its superproperty prov:wasDerivedFrom is a bit too weak, but
>>>>> can be used here.
>>>>>
>>>>> So if there was a prov:hadSource between these two, I would gladly use
>>>>> it, not having the strong requirements for the source.
>>>>> Otherwise I will settle for prov:wasDerivedFrom for now.
>>>>>
>>>>> Thank you again, that helped a lot!
>>>>>
>>>>> Cheers,
>>>>> Denny
>>>>>
>>>>>
>>>>> 2012/8/8 Jun Zhao <jun.zhao@zoo.ox.ac.uk>:
>>>>>
>>>>>> Hi Denny,
>>>>>>
>>>>>>
>>>>>> On 08/08/2012 13:05, Denny Vrandečić wrote:
>>>>>>
>>>>>>>
>>>>>>> Every statement has an IRI. And the source will also have an IRI
>>>>>>> describing it (i.e. an IRI for the statistical yearbook, an IRI for
>>>>>>> the mentioned paper).
>>>>>>>
>>>>>>
>>>>>>
>>>>>> By this, do you really specially asking for a property to express
>>>>>> that a
>>>>>> statement about an entity is derived from a certain primary source?
>>>>>>
>>>>>> How about the http://www.w3.org/TR/prov-o/#**hadPrimarySource<http://www.w3.org/TR/prov-o/#hadPrimarySource>
>>>>>> ?
>>>>>>
>>>>>> Note that there are some renaming going on around this property. If
>>>>>> you want
>>>>>> to use more than the above property, such as prov:source or
>>>>>> prov:qualifiedSource, then please ping us before you do so.
>>>>>>
>>>>>> HTH,
>>>>>>
>>>>>> Jun
>>>>>>
>>>>>>
>>>>>>
>>>>>>> What I did not figure out is: which property from the provenance
>>>>>>> ontology can I use to connect the statement IRI to the source IRI?
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Jun Zhao, PhD
>>>>>> EPSRC Postdoctoral Fellow
>>>>>> Department of Zoology
>>>>>> University of Oxford
>>>>>> Tinbergen Building, South Parks Road
>>>>>> Oxford, OX1 3PS, UK
>>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
> --
> Jun Zhao, PhD
> EPSRC Postdoctoral Fellow
> Department of Zoology
> University of Oxford
> Tinbergen Building, South Parks Road
> Oxford, OX1 3PS, UK
>