remove subsumption axiom `dcat:Dataset rdfs:subClassOf dctype:Dataset.`

[dr-shorthair]

dctype:Dataset is part of the DCMI Type Vocabulary, alongside text, image, sound, software and some other genre terms.

In DCAT 1.0 dcat:Dataset is a subclass of dctype:Dataset. However, the scope of DCAT 'dataset' is more general than the DCMI Type 'dataset' (see Scope or type of dataset with a DCAT description [ID8]), and also includes 'text' at least, and maybe multimedia types as well - the Datacite-DC-DCAT mapping provided at #64 (comment) certainly shows this.

So, while the DCMI Type Vocabulary could reasonably used as a typing vocabulary for classifying a dcat:Dataset (see #64), it is incorrect to pre-empt this by making dcat:Dataset a subclass of dctype:Dataset.

[makxdekkers]

I agree that it is too restrictive to declare dcat:Dataset a subclass of dctype:Dataset.

[dr-shorthair]

Resolved in meeting https://www.w3.org/2018/02/07-dxwgdcat-minutes
Implemented in #108

[agbeltran]

Re-opening until the PR is merged.

[dr-shorthair]

@agbeltran action on the PR is currently with you.
Do you see any problems?

[agbeltran]

@dr-shorthair the PR is actually about two issues (I suggest we do a PR per issue, what do you think)? I'm still reviewing one of the issues.

[dr-shorthair]

This issue is fully resolved by #108 so can be closed.
#64 should remain open until documentation is completed, per the comments.

I combined two changes in one PR as they are logically bound together - the subsumption axiom is replaced by (... generalised to ...) the dct:type property.