IsServiceFor and IsServedBy in DataCite

[gilles landais]

Hello Markus,

We have already had that discussion in DCP.

That is true, that there is no equivalent term in DataCite. Using the 
same semantic is clearly better for relations that exists in both 
schemas. But I'm not sure that we need a complete bijection.
The DataCite and registry metadata are different because the aim are 
also different.

I don't know if DataCite will accept - If they accept, that' fine! But 
I'm skeptical ...
Is it the role of DataCite to provide the same level of interoperability 
than the registry?

... DataCite focuses on citation and discoverability (with relationship 
between resources) but not on how querying the datasets. That's not the 
same idea of interoperability...


(if DataCite won't accept) there are other schema that allows to make 
the links dataset-service.
I think about DCAT which is a W3C used in OpenData. Investigate this way 
could improve the discoverability in other networks.
(I make some test with VizieR records provided through OAIPMH, but that 
's not ready yet)


Regards

Gilles


Le 30/10/2025 à 11:48, Markus Demleitner via registry a écrit :
> Dear Registry folks,
>
> As part of our scheme for discovering data collections
> <http://ivoa.net/documents/discovercollections/>, we need
> relationship types IsServedBy and IsServiceFor in our relationship
> type vocabulary
> <https://www.ivoa.net/rdf/voresource/relationship_type/>.  Actually,
> these concepts date back to vintage 2006 VOResource 2.0 (served-by
> and service-for).
>
> When we moved to DataCite relationships and the RDF vocabulary in
> VOResource 1.1, I freely added these two terms in DataCite style, but
> without bothering to upstream them.  Now that we rather routinely go
> to DataCite metadata (mainly for DOI registration) and sometimes
> back, too, it is inconvenient that these relationships cannot be
> represented in DataCite.
>
> I had hoped that other communities would feel the same itch and
> DataCite would grow these things itself.  But our sort of service
> apparently still hasn't quite arrived in too many places out there.
> So, I went ahead and wrote a suggestion myself:
>
> https://github.com/datacite/datacite-suggestions/discussions/207
>
> As you can see, there are many suggestions for the DataCite people to
> consider; I'm hence somewhat skeptical that my puny suggestion alone
> will be enough to get the relevant working groups moving.  So, if you
> could contribute to the suggestion or perhaps know people involved
> with DataCite that could be helpful for moving this ahead: Any help
> is much appreciated.
>
> Thanks,
>
>            Markus
>
>
>
>
>