Add IsServiceFor and IsServedBy to relationshipType

[msdemlei]

What is the problem that your suggestion solves?

In the Virtual Observatory, the notion of a service (a standard interface offering access to data collections) is very important; for instance, a single Table Access Protocol service might offer access to thousands of tables, each of which is a resource (in the datacite sense) in itself; to make it actionable, however, a declared relationship to the service is necessary.

In our Registry, we have been moving towards adopting DataCite relationship types (cf. https://www.ivoa.net/rdf/voresource/relationship_type/); in converting between the VO and the DataCite metadata schemas, it is regularly a problem that DataCite does not allow one to declare relationships of this kind.

What solution might meet your needs?

It would be wonderful if DataCite could adopt our additional concepts:

IsServiceFor -- This resource makes data from the referenced resource (typcally some sort for data collection) available (inverse of IsServedBy)

IsServedBy -- This resource can be accessed or otherwise used through the referenced service (inverse of IsServiceFor)

Your name

Markus Demleitner

Your organization

University of Heidelberg

What alternatives have you tried or considered?

I don't think there are any plausible ones to address the use case. The alternative we are using right now, hiding these relationships from DataCite/DOI and mentioning the access mode on the landing pages, is not very attractive; it involves special handling by relationship type, and it turns machine-readable information into non-machine-readable information, which is rarely a good thing.

Is there anything else you would like to share?

No response

What group(s) would benefit from your suggestion?

• DataCite members registering DOIs and metadata and administering accounts
• Developers building integrations with DataCite services
• Harvesters who index DataCite DOIs and metadata
• Individuals or institutions analyzing DataCite DOIs and metadata
• DataCite staff
• Other

If other group(s), please describe.

The Virtual Observatory community (https://www.ivoa.net)

[acraugh]

This is similar to a use case we have in the Planetary Data System - specifically, the CATCH interface (https://catch.astro.umd.edu) that enables users to search a specific collection of sky surveys to retrieve images and data across a collection of sky surveys for a specific target of interest. It would also be potentially applicable to dataset-specific services elsewhere in the PDS, like the OPUS (https://opus.pds-rings.seti.org/opus/#) service or the Analyst's Notebook (https://an.rsl.wustl.edu/) service.

[KellyStathis]

Thanks for this suggestion @msdemlei and for your use case @acraugh! This is very similar to this suggestion: #34 (which I think came from @acraugh originally, before we started using GitHub Discussions).

One thing I am wondering about is the connection between a "service" in the sense you're using it here and the concept of a "Hosting Institution", which is one of our contributor types: https://datacite-metadata-schema.readthedocs.io/en/4.6/appendices/appendix-1/contributorType/#hostinginstitution. I know that the service isn't itself an organization, but do you think the link between a "service" and the institution that supports it would be relevant to include?

[msdemlei]

On Mon, Nov 03, 2025 at 10:31:57AM -0800, Kelly Stathis wrote: One thing I am wondering about is the connection between a "service" in the sense you're using it here and the concept of a "Hosting Institution", which is one of our contributor types: https://datacite-metadata-schema.readthedocs.io/en/4.6/appendices/appendix-1/contributorType/#hostinginstitution. I know that the service isn't itself an organization, but do you think the link between a "service" and the institution that supports it would be relevant to include?

In VOResource <https://ivoa.net/documents/VOResource/>, the institution is modelled as the publisher of a service, which I think has served us well. Do you remember what scenario it was that prompted the split between "Publisher" and "HostingInstitution" in DataCite? In the Virtual Observatory, I can't think of an instance in which these would be distinct, but that may be a domain specificity. Anyway, I think it is safe to say that a HostingInstitution would be identical to the Publisher in the case of what we call a service here (roughly, a set of resources with a defined machine-readable representation).

[acraugh]

Interesting question, @KellyStathis. In the PDS case, the HostingInstitution of the service would be different from the Publisher of a data set. It would usually - but not always - be the same as the Curator of the data set.

The Publisher of most of our data sets is a national space agency - NASA, ESA, JAXA, etc. The service might access a local copy of the data curated by the institution running the service, or it might access a remote copy of the data hosted by another institution. I can go into details if you really want to know them, but the key is that these national space agencies have agreed on a single metadata standard in order to promote the development of interoperable services that do not require local copies of data in order to be able to search for and access data across the national planetary archives.

We think of the PDS nodes as libraries, rather than publishers.