Faceted search is an attractive and user friendly information retrieval paradigm. It combines two traditional approaches, namely keyword search (allowing the direct search) and navigational search (allowing for browsing the information space by iteratively narrowing the scope). In creating a faceted query, the user is provided with a tree-shaped faceted interface, which is a hierarchical view of an ontology describing the information domain. In this paper, we discuss a transformation and translation process leading from a graphical form of a faceted query, through first order forms (without and with rewriting) to a SQL query. The crucial in this is the mapping of the ontology into a relational database, and translating to SQL. In DAFO system we have used a commercial RDBMS with optimization capabilities. We report our experimental results which show very promising execution times.