The advantage to this is that it works as an ad-hoc query language for existing graph databases. I could use Gremlin to generate views of the data, and have oth

The advantage to this is that it works as an ad-hoc query language for existing graph databases. I could use Gremlin to generate views of the data, and have other programs operating on database independently. Prolog doesn't really play well with others -- it's effectively its own operating system like Smalltalk or Pick.