Hoi,
The good news is that the interwiki tangles are a big improvement over what we had. Now there is clarity when an interwiki is wrong. There is one place where to solve it and when solved, it is solved for all linked Wikipedias.

The problem I have with all the ontology issues is that they are relevant in the context of interconnectivity between systems. This becomes increasingly irrelevant as we identify the records in other systems. Irrelevant it is also because we do not use it. Getting data from external sources is largely frowned upon.

As it is, we are building the data in Wikidata ourselves and comparing things is an afterthought. Many statements in Wikidata are problematic but they have the saving grace that with more statements it becomes more clear what works and what does not. As long as people talk in terms of ontologies and do not translate it into "Reasonator" like application it is in my opinion a waste of breath.
Thanks,
       GerardM

http://ultimategerardm.blogspot.nl/2013/10/abdullah-king-of-saoudi-arabia.html
http://tools.wmflabs.org/reasonator/?q=Q57298


On 6 November 2013 08:24, Jane Darnell <jane023@gmail.com> wrote:
Hi TomTOm,
Be careful what you wish for! If this were possible, then if someone changed the dates, this could mess up other things. We already have a big job untangling mismatched interwiki links, and this would make such mismatches possible to the nth degree.
Jane

Sent from my iPad

On Nov 4, 2013, at 6:14 PM, Thomas Douillard <thomas.douillard@gmail.com> wrote:

>
> Hey, I got an ontology question.
>
> Classes are, in semantic web framework and their foundations like Description Logic, if I'am not wrong, something like a lohic predicate that intensionaly or extentionaly defines the properties of their instances.
>
> They are usually not qualified, but in Wikidata, as of now they are properties like the others, who can also be qualified.
>
> So the question is : could we use qualifier on classes to add predicates on the class definition ? For example if
> <George Bush> is an instance of <United States President> [<from> 1980 <to> 1984] (random years), this would mean that the instanciation add some predicates on the other predicates we have on the <president of the united states> ?
>
> Just a random thought, I just realise I just qualified the instanciation, not the class itself.
>
> --TomT0m
>
> _______________________________________________
> Wikidata-l mailing list
> Wikidata-l@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikidata-l

_______________________________________________
Wikidata-l mailing list
Wikidata-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata-l