WIREFRAME: two-phase, cost-based optimization for conjunctive regular path queries
Parke Godfrey, Nikolay Yakovets, Zahid Abul-Basher, Mark Chignell · TU/e Research Portal · 2017
While graph databases and their applications are coming into wide use, we are only at the very beginning of understanding how to scale these systems well. Recent work has brought a cost-based optimization approach to RPQs [6]. We set out a framework herein which we call Wireframe for a two-phase, cost-based optimization for CRPQs. In Wireframe, CRPQ planning—and, likewise, evalu- ation—is separated into two phases. In the first phase, the plan is for evaluating the “answer graph”. In the second phase, a plan is posited for enumerating the subgraph-match answer tuples from this answer graph.