There's little need. It's a direct application of concepts that are well treated academically in all the various datalog papers.
But also, have you read any of the papers involved? Datalog is pretty simple. It's a restricted, forward chaining prolog. Once you know that, you can recreate most of it from that description alone.