Interactive Pay as You Go Relational-to-Ontology Mapping

摘要

Ontology Based Data Access (OBDA) enables access to relational data with a complex structure through ontologies as conceptual domain models. To this end, mappings are required. A key aim of OBDA is to facilitate access to data with a complex structure. Ironically, though, in today's existing OBDA systems mappings typically need to be compiled by hand, which is a complex and labor intensive task. Additionally, existing semi-automatic mapping approaches suffer from high human effort for cleaning up results. Fully automatic approaches, on the other side, suffer from a lack of precision and/or recall. In setups where the correctness of query results is crucial but the initial human effort must still be kept be small as possible, neither approach is acceptable. This situation calls for a guided, pay as you go feedback process for human mapping validation. We envision a comprehensive suite of methods and techniques that work well with one another in a seamless mapping process and support mapping construction in the context of OBDA. This suite will in part consist of a recombination and adaptation of various existing methods, but will also comprise newly devised algorithms and techniques.