Question about hasCreationDate

[Sylvain Loiseau]

Dear all,

A seemingly simple question, but I’m having trouble finding a clear answer: is
https://www.ica.org/standards/RiC/RiC-O_1-1.html#hasCreationDate intended to express the date on which a resource (such as a Record) is created, or the date on which the material Record being described was created? I am inclined to think it is the latter, but I would appreciate confirmation.

If that is indeed the case, it raises a related question: how should one record the history of the description itself (e.g. when an entity was created or last updated), assuming this is something that needs to be tracked?

Best regards,

Sylvain

[Florence Clavaud]

Dear Sylvain

As specified in RiC-O (and in RiC-CM), when used, rico:hasCreationDate connects a Record Resource (as the message conveyed) or an Instantiation (as the inscription of this message on a carrier) to the Date when it was created. An Instantiation may have its own creation date, e.g. when it results from the digitization of an analogue one.

A Record Resource may describe another Record Resource (in which case it can be connected to the described Record Resource using rico:describesOrDescribed). RiC considers descriptions of record resources (or any other entity) as records. Therefore, such descriptions may have the same attributes or the same relations as any other Record Resource. RiC-O also defines two individuals, AuthorityRecord and FindingAid, as instances of both skos:Concept and rico:DocumentaryFormType, to be used when needed (see https://www.ica.org/standards/RiC/RiC-O_1-1.html#individuals).

See on this topic section 6 (Documenting Description) of RiC-CM 1.0, pp. 125-129. 

Many examples of such Records are available here: https://github.com/ICA-EGAD/RiC-O/tree/master/examples/examples_v1-1/NationalArchivesOfFrance/rdf-xml/recordResources, and in https://github.com/ICA-EGAD/RiC-O/tree/master/examples/examples_v1-1/NationalArchivesOfFrance/rdf-xml/agents (download and open any of the files stored there).

Hope it helps,

Florence Clavaud
chair of ICA/EGAD, lead of RiC-O development team

[Johan Pieterse]

Hi

My opinion

This is a question that surfaces a genuine ambiguity in RiC-O, and one that has direct practical implications for implementations.

On hasCreationDate

I believe your inclination is correct.

ric:hasCreationDate is a property of the described resource (the Record, RecordSet, Instantiation, etc.), not of the RDF description itself. In RiC-CM terms, it corresponds to the "Dates of Creation" attribute of a Record entity, the date(s) the intellectual content or physical instantiation came into existence. The ontology is modelling the archival object, not the metadata act.

On description history, the RDF/OWL-native answer:

The standard mechanism for this in RDF is dcterms:created and dcterms:modified on the named graph or the description resource itself. If you are using named graphs, you can assert:

<https://example.org/graph/record-001> dcterms:created "2024-01-15"^^xsd:date ;

    dcterms:modified "2025-03-10"^^xsd:date ;

    dcterms:creator <https://example.org/agent/archivist-jane> .

RiC-O does not currently define properties for provenance of the description, it deliberately stays in the domain of the described entities. This is a known gap, and PROV-O (prov:wasGeneratedBy, prov:wasAttributedTo) is the most principled complement for description provenance if you need full auditability.

A practical note

If you are tracking this in a system like AtoM (I developed RIC pulling directly from AtoM) or a triplestore, the pattern I suggest is to treat the description as itself an entity, either a prov:Entity or a rico:Record describing a rico:Record, and apply creation/modification dates at that level. Somewhat recursive, but ontologically sound.

Johan Pieterse
082 337-1406

--
You received this message because you are subscribed to the Google Groups "Records_in_Contexts_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to Records_in_Context...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/Records_in_Contexts_users/5f26d4ba-0205-4d5a-aa13-d58f1ded4b2an%40googlegroups.com.

[Richard Williamson]

Dear Johan,

Thank you very much for your thoughts! I guess that your message may
have cross-posted with Florence's earlier this week; as Florence
wrote, the RiC point of view is to regard an archival description
itself as a Record, in the manner you suggest at the end of your
message. In addition to what Florence wrote, there is a little on this
in RiC-AG in §6a.3.4, particularly the second example there:

https://ica-egad.github.io/RiC-AG/faq--general_questions_and_smaller_modelling_questions.html#do-you-have-examples-illustrating-chapter-6-of-ric-cm

In particular, I would not say that it is a 'known gap' conceptually
:-). On the other hand, I think it is definitely true, as indeed has
been discussed previously on the list, that there is room for
'shortcuts' at least in RiC-O for this kind of situation, and also for
mappings e.g. to PROV-O. One complicating factor is that there is not
really one single best practice for this in RDF generally; there are
things like RDF-Star for instance, closely related to the kind of
things you mention in the middle of your message. Even more so, there
is very little that is canonical on this front in the OWL world. It
may therefore be that we have to step outside of OWL/RiC-O itself if
something along these lines is to be provided. But what can at least
be said is that the RiC-O team within EGAD is aware of this kind of
direction, and actively considering what can be done; as I say,
though, I see these concerns as primarily 'syntactic', semantically
one can express it already in a somewhat more verbose way :-).

Best wishes,
Richard

> To view this discussion visit https://groups.google.com/d/msgid/Records_in_Contexts_users/CAA4s_cBnEab6v79f_vqJhdOzup%2BM%2B_kzBa-%3D%3DvE-yoNCk_%3DM4w%40mail.gmail.com.