|Last release:||1.7 (2012/04/17)|
|License:||Free for non-commercial use|
|Author(s):||Mariano Rodriguez-Muro, Diego Calvanese, Josef Hardi|
|Affiliation:||Free University of Bozen-Bolzano|
Quest is a reasoner focused on fast and efficient SPARQL query answering over RDFS and OWL 2 QL inference regimes and in the context of Ontologies with mappings to relational data sources (OBDA). The core of Quest is based on query-rewriting, however, it is highly optimized to avoid common issues with this reasoning technique.
Features included in Quest are:
- Support for RDF(S)/OWL2QL inference regimes
- SPARQL query answering
- Very flexible, expressive and easy to learn datasource mapping language
- Support for Virtual ABoxes or virtual RDF graphs modes, in which the data is never imported into the reasoner and remains in the original source which is accessed on-the-fly
- Support for any JDBC data source, including database federation (a.k.a. database virtualization) systems such as Teiid
- Capable of handling billions of Data Triples (a.k.a. ABox assertions)
- Support for OWLAPI 3
- Available as a plugin for Protege 4.1
The current implementation of Quest is able to handle ontologies with thousands of Classes/Properties and thousands of RDFS/OWL2QL axioms. Ontologies with under 2000 axioms, and less than 1000 classes/properties can be loaded by quest in less than 30 seconds. When working in classic OBDA mode, the number of data triples (ABox assertions) is only limited by the amount of RAM available to Quest. In Virtual OBDA mode there is no limit in the number of triples/ABox assertions; since in virtual mode Quest leaves all data in the source, the size of the data source doesn’t affect Quest’s startup time. Expect the system’s performance to increase with each new release, as we improve our implementations.