Reinforcement Learning-based SPARQL Join Ordering Optimizer

In Proceedings of the 20th Extended Semantic Web Conference: Posters and Demos (2023)

In recent years, relational databases successfully leverage reinforcement learning to optimize query plans. For graph databases and RDF quad stores, such research has been limited, so there is a need to understand the impact of reinforcement learning techniques. We explore a reinforcement learning-based join plan optimizer that we design specifically for optimizing join plans during SPARQL query planning. This paper presents key aspects of this method and highlights open research problems. We argue that while we can reuse aspects of relational database optimization, SPARQL query optimization presents unique challenges not encountered in relational databases. Nevertheless, initial benchmarks show promising results that warrant further exploration.