Have been looking at Omeka S Sandbox, and one thing that jumped out is the incorrect use of the term “vocabulary”. The Vocabularies section (under Resources) actually refers to Metadata Schemas, not Vocabularies.
“Vocabs/vocabularies” is used correctly in other contexts, e.g. the “Custom Vocab” and “Value Suggest” modules.
The terminology is all a bit new for Omeka and as always there’s some overlap, but it’s my understanding that “vocabulary” isn’t uncommon at least in the way its being used there: see e.g. Linked Open Vocabularies, or the World Wide Web Consortium’s linked data glossary.
Or am I misunderstanding your issue with the usage?
Hi John, I’m a librarian, and “vocabularies” are used within this profession to refer to lists of preferred terminology e.g. Library of Congress Subject Headings, the Boolean yes/no, Getty Thesaurus of Geographic Names etc.
My judgement is that the way it is being used in your 2 example links is incorrect. The “vocabularies” referred to in those links are actually metadata schemas, as they describe the fields which make up the schema/resulting templates. If they were describing terms that could go into the fields, they could correctly be called vocabularies.
This sounds like a problem of Omeka S being used in a variety of contexts and disciplines. Some of the terminology does clash across those disciplines. If we keep to the usage in one discipline, we’ll lose the usage in another discipline, so it will always be a nuance of context. Since Linked Open Data folks use ‘Vocabulary’ often enough, and Omeka S is oriented around LOD, it seems best to keep to that terminology while we’re in the context of those metadata schemas / vocabularies.
Hi Patrick, problem here is that the LOD folks can use “Vocabulary” without any confusion, as they’re only using it in the context of a single meaning (metadata schemas). But Omeka S is using it in both its meanings, viz. for the Resources - Vocabularies (metadata schemas) and for controlled vocabularies (aka the ‘Custom Vocab’ and ‘Value Suggest’ modules.
This is highly likely to cause confusion.
In RDF standard, vocabulary has the same meaning as ontology. This concept is a specific use of controlled vocabularies.
The Dublin Core is a RDF ontology (or RDF vocabulary, it’s the same). Since it describes objects, it’s also a controlled vocabulary.
In fact, all RDF vocabularies are controlled vocabularies (and I think all controlled vocabularies can be standardized in RDF way). I don’t see how it can cause confusion.
We’ve struggled with the overlap and different usages, but haven’t found an alternative that doesn’t itself introduce more confusion. We’d end up using terminology that is unfamiliar to everyone, so we’re just going with commonly used terms in their subtle contexts.
I’m not quite sure that RDF vocabulary is the same as a controlled vocabulary in the way bernieh describes it. RDF vocab/ontology describes the properties, not property values – the controlled vocabularies.
That’s what leaves me saying that, while it is complex and nuanced when we get into the details, having them exist together with a switch based on context seems unavoidable.