Multiple labels in Vocabularies


I have an ontology built with multiple labels for each class and multiple comments, depending on language.
I would like to import the multiple labels and differentiate resource templates for language. I know and have already used the mechanism of alternate labels and comments in Resource templates.
But I would like to exploit the multiple label mechanism to provide different labels as required. Without generating alternate labels that are not in the already published ontology or replicating them by hand.
Is that possible within the platform or should I necessarily develop an external website relying directly on ontology (and introducing a double access to the ontology)?
Alternatively, I can upload the ontology on two different public repositories (e.g., and, respectively) and import two vocabularies that are actually the same. Not a great solution.

Thank you
Vincenzo Lombardo

If I’m understanding correctly, you have a defined vocabulary that has multiple labels in different languages and you’d like to use those without having to re-write them manually?

The vocab member labels get handled through the translation component, so that might be a more natural place for a solution here than the template/alternate label system. We don’t really have any convenient system for adding in translations from vocabularies, though… it can be done with a module for example, but you need to provide the translations.

In your ideal scenario, what does this look like? You’d like to read the labels for other languages out of the vocabulary’s RDF definition and store/use them?


Thanks for your reply.

Your description of the ideal scenario is perfect. And I believe it is ideal for the relationship with RDF vocabularies, which admit an indefinite number of labels with a single class or property.
One should be able to import multiple labels and use them as preferred. In alternative, one could admit, several vocabularies with the same Namespace URI, with import pivoted on labels.
But maybe I am not identifying some incorrect implications.

Thanks again for your support.

So, as I mentioned before, the translation system that translates the user interface strings and things like that also can handle property/class names. For the default ones we have those strings built into the translation sources our volunteer translators work on, so we can just ship those with Omeka. But for a custom vocabulary, we can’t do that obviously.

I’m thinking one option that could work without too much trouble and would work with current versions of Omeka S would be a module. Modules can add translations, so you could have a module that takes the RDF document for a vocabulary and reads out all the translated labels and sets those in the translation system.

In that kind of setup, the labels would just follow the installation’s/site’s/user’s setting for language. Would that work well for your purposes?

A module that adds to the translation system (as it works now for the Dublin core terms, I assume) should work well.

OK, good to know.

Do you have an example of an existing RDF vocabulary document that is set up in this way that you’d like to be supported?

Sure. Here it is:
Bilingual labels in English and Japanese
Currently used in the BeArchaeo project, where we introduced alternative bilingual labels: