I'm attempting to create/implement an RDF-schema for historical data: setting up a base schema is simple, but I've run into a roadblock when trying to add a verifiable source for each claim (predicate).
For example, it's relatively easy to declare 'Alexander (subject) has a brother (predicate) Robert (object)'... but I would like that claim (that is the whole) that Robert was indeed Alexander's brother to be cited, with a link to the original source document, if possible.
The best I could come up with is a 'doubled triplet' schema:
Alexander -> has_a_brother -> Robert
Alexander -> has_a_brother_source -> 'http://sourcedocumentserver.com/blablabla'
Alexander -> has_a_brother_source -> 'http://sourcedocumentserver2.com/blablabla'
Alexander -> has_a_brother_source -> 'http://sourcedocumentserver3.com/blablabla'
Is there a better way to go about this?
PS: this problem is a bit different than a 'predicate variable' (like the question that this one was said to replicate), but in this 'citation needed' circumstance, every single predicate in the database would need a secondary description/label/name (and that's even without imagining multiple citations, attribution data, etc.).
The best I could find was transforming every 'citation needed' predicate into a named graph (with multiple RDF triplets) that would contain the object normally directly joined (if no citation were required).