This implementation assumes there is no “document repository” corresponding to a “document” entity. So, if 3 finding aids are related to same document, it’s described 3 times within descriptions. I would prefer to develop entity/relations between descriptions and documents. This orientation implies IT systems offer a “document repository” such as existing one for “agents” or “places”. It would normalize and mutualise “document” description as an entity.
In this repository, we may describe document as entity, which means:
A document is a conceptual view of archival document
Which is materialized by (1..n) implementation(s) (instanciations)
Document is described by some hig-level metadata : persistent identifier, creation date, destruction date, creator, type, description, volume, dimensions, damages, history, type (photo, map, document, audiovisual, etc.), etc.
An implementation is described by local metadata : materiality (born digital, digitized, physical), usage (preservation/dissemination/etc.), parent (“derived from…”), traceability note, creation date, persistent identifier, etc.
In this way, a digitized implementation of one physical document may gather several files (1 tiff file = 1 view/ page)… ordered within a “intellectual plan” that might be a IIIF manifest. So, implementation corresponding to digitized document… may group several files (hierarchical structure).
Possibly we can add “technical” metadata to implementation : file format (pronom), hash, size (bits), etc.
Relations between implementations, with finding aids, with agents or places.
You’ll find in attachment a draft modelling of “archive” entity that may be linked to finding aid. This entity represents intellectual document and attach all its known instances. So, entity/relation may be fully exploited in EAD. This draft needs to be improved but give an idea of data model we could promote for descriptions.
Creator of issue
The issue relates to
Wanted change/feature
You’ll find in attachment a draft modelling of “archive” entity that may be linked to finding aid. This entity represents intellectual document and attach all its known instances. So, entity/relation may be fully exploited in EAD. This draft needs to be improved but give an idea of data model we could promote for descriptions.
Diagram.pdf