Ah ok, so in RDF terms each predicate gets its own table, which makes sense. Then you have a subject foreign key relation ship and a object foreign key relation ship to the node tables.
That would be better than one big table in performance, which was a major problem in the RDF on SQL databases.
Of course those accepted any graph, if one constrains the number of possible predicates/relations then this solution could more efficient.
Comments
Ah ok, so in RDF terms each predicate gets its own table, which makes sense. Then you have a subject foreign key relation ship and a object foreign key relation ship to the node tables.
That would be better than one big table in performance, which was a major problem in the RDF on SQL databases.
Of course those accepted any graph, if one constrains the number of possible predicates/relations then this solution could more efficient.