Abstract
Controlled languages (CLs) are ambiguity-free subsets of natural languages such as English offering a good trade-off between the formal rigor of ontology and query languages and the intuitive appeal of natural language. They compositionally map (modulo a compositional translation τ (·)) into (or express) formal query languages and ontology languages. Modulo compositionality, they inherit the computational properties of such ontology/query languages. In the setting of OBDAS, we are interested in capturing query answering and measuring computational complexity w.r.t. the data queried (a.k.a. data complexity). In this paper we focus in defining a CL capable of expressing a subset SQL aggregate queries, and study its data complexity w.r.t. several ontology languages and extensions of the query language.