Query Answering over Complete Data with Conceptual Constraints

Abstract

Query answering over databases with conceptual constraints is an important problem in database theory. To deal with the problem, the ontology-based data access approach uses ontologies to capture both constraints and databases. In this approach, databases are considered under open-world assumption which creates many issues including the necessity of restricting to only positive queries, and the failure of query composition. In our research, we focus on a combined approach that allows data in databases stays completely as under closed-world assumption while knowledge providing by conceptual constraints can be incomplete. We first study the complexity of query answering problem under description logic constraints in the presence of complete data and show that complete data makes query answering become harder than query answering over incomplete data only. We then provide a query rewriting technique that supports deciding the existence of a safe-range first-order equivalent reformulation of a query in terms of the database schema, and if so, it provides an effective approach to construct the reformulation. Since the reformulation is a safe-range formula, it is effectively executable as an SQL query. At the end, we study the definability abduction problem which aims to characterize the least committing extensions of conceptual constraints to gain the exact rewritable of queries. We also apply this idea to data exchange - where we want to characterize the case of lossless transformations of data.