Agreed. Actually even the openlink virtuoso database project is kind of using the RDBMS approach in its underlying implementation. https://en.wikipedia.org/wiki

Agreed. Actually even the openlink virtuoso database project is kind of using the RDBMS approach in its underlying implementation. https://en.wikipedia.org/wiki/Virtuoso_Universal_Server https://en.wikipedia.org/wiki/Virtuoso_Universal_Server However part of the power of the graph database that utilising a linked dataset such as RDFs is that you will be able to use SPARQL to query the graphs, which is more suitable for sets / graphs.