search

opendata-swiss / dcat_ap_ch

Examples for geocat and DCAT data-catalogs are given here
5 stars 3 forks source link

dct:coverage on dcat:Distribution does not conform to DCAT-AP #79

Open sabinem opened 3 years ago

sabinem commented 3 years ago

Property: dct:coverage Class:dcat:Dataset and dcat:Distribtution Conformance Problem:

Proposal:

Changed Proposal: Watch what happens on DCAT-AP and DCAT regarding the modeling on Series of Data, since this seems to be a topic where there is some movement in the DCAT community.

Juan-Juan-1 commented 2 years ago

Isn't it a opendata.swiss issue too...? What would be the consequence of removing this property? I am convinced it should be removed, but we should also think about the consequences..

metaodi commented 2 years ago

The reason that dct:coverage exists on the dataset is the lack of support for polygons in [dct:spatial(https://dcat-ap.ch/releases/2.0/dcat-ap-ch.html#dataset-coverage):

It is not possible with dct:Location to provide a polygon as geographical coverage, therefore dct:coverage should be used for this purpose (§ 7.3.14 Property: coverage)

So this part:

In this case the DCAT properties dct:spatial and dct:temporal are able to completely replace dct:coverage on dct:Dataset.

is not true.

Of course we could decide that the range of dct:Location (that dct:spatial provides) is enough and arbitrary polygons are not necessary. But I think it's important to have the full picture why this property was added in the first place.

metaodi commented 2 years ago

In #61 it is stated that dct:spatial now supports polygons (via GeoDCAT). If this is the case, forget my previous comment and get rid of dct:coverage 😁

andreasamsler commented 2 years ago

I would need an example of a Dataset/Ressource at hand to be able to understand from a user's pespective what need(s) we want to cover with this change.

Juan-Juan-1 commented 2 years ago

Hi @andreasamsler, in this case there wouldn't be a big change for the user. The catalogue would be structured slighlty differently, so that the information may (it depends how opendata.swiss would implement this change) be presented in slightly in a different way. It's way more a compatibility issue with DCAT-AP: this value does not exist there and this kind of information is carried through dct:spatial and dct:temporal instead. This may lead to issues like the information is not made available on data.europa.eu.

(Note that I'm not taking into account the topic "dcat:Distributions Usage note should not encourage the usage of Distributions as Dataseries #112", since it's only indirectly linked to it).

sabinem commented 2 years ago

@metaodi and @andreasamsler I have launched a discussion about this topic of modeling a Series as Distributions on DCAT-W3C: the suggestions there are quite good. Feel free to jump right into that discussion: https://github.com/w3c/dxwg/issues/1429

For dct:coverage on dcat:Dataset I created a new issue, since this case is more clear and can be decided already, whereas the case of dct:coverage on dcat:Distribution is harder to solve until DCAT and DCAT-AP over support for the related use cases.

metaodi commented 2 weeks ago

I think we should include this proposal in 3.0.0 to a) make clear dct:coverage should not be used for DatasetSeries b) make sure #214 is adopted as well, which actually add the new DatasetSeries class

metaodi commented 2 weeks ago

Update: I think we should close this issue and use the proposal in #112 instead.