A Proposal for Deploying Hybrid Knowledge Bases: the ADOxx-to-GraphDB Interoperability Case

Abstract

Graph Database Management Systems brought data model abstractions closer to how humans are used to handle knowledge - i.e., driven by inferences across complex relationship networks rather than by encapsulating tuples under rigid schemata. Another discipline that commonly employs graph-like structures is diagrammatic Conceptual Modeling, where intuitive, graphical means of explicating knowledge are systematically studied and formalized. Considering the common ground of graph databases, the paper proposes an integration of OWL ontologies with diagrammatic representations as enabled by the ADOxx metamodeling platform. The proposal is based on the RDF-semantics variant of OWL and leads to a particular type of hybrid knowledge bases hosted, for proof-of-concept purposes, by the GraphDB system due to its inferencing capabilities. The approach aims for complementarity and integration, providing agile diagrammatic means of creating semantic networks that are amenable to ontology-based reasoning.