In this architecture modelers
create OBDA models that include i) The Ontology ii) The
OBDA mappings and iii) access information for the data
sources. An OBDA-enabled reasoner is provided with this
information and it is trough this reasoner that clients
(i.e., users or applications), have access to the data from
the sources. The users are only exposed to the Ontology,
which provides a clean, semantically rich and homogeneous
views of the source(s).
Key to this picture are the availability of tools for
describing OBDA models, this is where the OBDA Plugin comes
in. The OBDA Plugin is an add-on for the
Protege ontology editor aimed at transforming Protege into
a fully fledged OBDA model editor. It
provides data source and mapping editors, as well as
querying facilities that, in conjunction with an
OBDA-enabled reasoner compatible with the plugin, allows
you to design and test every aspect of an OBDA system. All
the data created by the modeler is stored in so called
.obda files, that, in conjunction with the OWL ontology,
can be used to deploy the OBDA-enabled reasoner in the
application that requires the semantic layer.
Sources are regarded as repositories over which queries can
be ran (considering a query can be any arbitrary
computation). Then mappings in the OBDA setting are viewed
as a pair of queries, the source query and the
ontology query. The definition of the semantics of
these the mappings is delegated to the OBDA-enabled
reasoner, this is done in order to allow for different
kinds of OBDA-enabled reasoners to use the plugins
facilities.
Currently, the OBDA plugin supports RDBMS sources and
mappings where the source query is an arbitrary SQL query,
and the target query is a conjunctive query over the
ontology with no existential variables. However, the plugin
is design with extensibility in mind and will soon allow
programmers to easily define their own kind of sources and
mappings. This way implementors of OBDA-enabled reasoners
can reuse the facilities offered by the plugin in their own
setting.
The OBDA plugin is based on the OBDALib, a parent project
intended to provide programmatic access to OBDA related
functionality. The OBDALib is not available at the moment,
but will be released soon, together with version 1.0 of the
OBDA plugin. At the moment, the OBDA plugin for Protege 4.0
is currently under heavy development and is only available
as a preview. Do not use it for production and contact
the authors for support.
Requirements:
- Protege 4.0 (Protege 4.1 and Protege 3.4 support will be available in the near future)
- An OWL reasoner implementing the OBDALib extensions for OBDA for query answering and reasoning
About the DIG compatible plugin: Development for the DIG 1.1, Protege 3.3 compatible version of the software (OBDA plugin for Protege and QuOnto's DIG Server) is suspended for the moment. All development is now focused on the OWLAPI/Protege 4 version of all the software. These versions are much more stable and provide many new features and performance enhancements. The packages are now under closed beta testing phase. If you would like to participate in the testing, please contact the authors.
About OBDA@FUB
For more OBDA resources at the Free University of Bozen Bolzano follow this link.
The OBDA Plugin is developed in the Knowledge Representation meets Databases (KRDB) research group at the Free University of Bozen Bolzano. It has been developed in the context of the European FET project Thinking ONtologiES (TONES) project.
Authors
- Ph.D. Mariano Rodríguez-Muro (homepage) (designer, programmer, project manager)
- Prof. Diego Calvanese (homepage) (advisor)
Contributors
- Manfred Gerstgrasser (programmer)
