| Many researchers have proposed using conventional relational databases to store and query large Semantic Web datasets. The most complex component of this approach is SPARQL-to-SQL query translation. Existing algorithms perform this translation using either bottom-up or top-down strategy and result in semantically equivalent but syntactically different relational queries. Do relational query optimizers always produce identical query execution plans for semantically equivalent bottom-up and top-down queries? Which of the two strategies yields faster SQL queries? To address these questions, this work studies bottom-up and top-down translations of SPARQL queries with nested optional graph patterns. This work presents: (1) A basic graph pattern translation algorithm that yields flat SQL queries, (2) A bottom-up nested optional graph pattern translation algorithm, (3) A top-down nested optional graph pattern translation algorithm, and (4) A performance study featuring SPARQL queries with nested optional graph patterns over RDF databases created in Oracle, DB2, and PostgreSQL. |
