Closed terjesyl closed 3 months ago
To the community is there any preference of any of these: what is the master, what is the derived?
If there is no specific reason for doing otherwise, I suggest following DCAT here.
I think we should consider use cases where also dcat:seriesMember are used as the master when dataset series are created independently from the datasets they contain, for example:
Hundreds of municipalities provide air quality datasets, all conforming to the same specification. The national environmental agency takes on the role of aggregator and creates a dataset series corresponding to all air quality measurments in the country as a whole for easy overview / access.
Note 1: the newly formed dataset series is expressed in a catalog and refers to datasets expressed in hundreds of different catalogs.
Note 2: in this scenario nothing prohibits that a dataset appears in multiple dataset series. (E.g. a regional dataset series in addition to a national.)
If we are to allow scenarios like this we cannot assume that dcat:inSeries is the master, if by that we mean that we mean that we do not consider it's inverse being present (for instance in portals).
Given that
dcat:next
anddcat:seriesMember
have been introduced, would it be worth adding a chapter similar to the DCAT chapter Use of inverse properties ? Especially important is the requirement that inverses "MAY be used only in addition to [its property], and that they MUST NOT be used to replace them."These are the properties I found in DCAT-AP with inverses.