I start up this question in a new thread as it goes into a different problem domain,
One of the option to model the object attributes is to store attributes in a separate relational table and use a SQL to join the graph query results. Now here is a situation, if we want to join, we need to keep some reference key in the RDF datastore, so that they can join together. But it comes out a issue, how we sync the data? Suppose we keep a user node in RDF with property user ID and we keep all other user info in relational table. When we do join, we can join by user ID. But if we delete that user record in relational table, most likely RDF will not delete that user node in graph datastore. Am I correct on this point or Oracle has some sort of way to maintain the data sync?