The verbalizer is a tool that generates natural language explanations in English for derivations in ontologies. The current prototype is restricted to an EL fragment of description logics. However, the tool is continuously being extended and improved. This software is based on Java and available under the Apache 2.0 license. More information is available on the project webpage: http://verbalizer.github.io
Protege plugin: verbalizerPluginProtege.zip
Sourcecode on github: github.com/marvinki/verbalizer
Required: Protégé 5 (http://protege.stanford.edu)
- Unzip the zip file
- Copy or move the jar file to the "plugins" folder of your Protégé installation (under Protégé 5, this is the folder Java/plugins)
- Classify an ontology (e.g. the enclosed example ontology)
- Navigate to an inferred axiom and click the question mark
- Choose "Proof Based Explanation"
We are currently using two collections of ontologies for evaluating the software. These are:
- The EL classification competition ontologies from ORE (OWL Reasoner Evaluation) 2015 ( ORE 2015 Reasoner Competition Dataset)
- The TONES repository (TONES Ontology Repository/Corpus). Since ontologies of different fragments of description logics are included, we use this collection to determine what proportion of of ontologies/inferred inferences are covered by our tool.
- Phone: +49 (0)731/50-24115
- Fax: +49 (0)731/50-24188
- Dr. Marvin Schiller
- Ulm University
- Institute of Artificial Intelligence
- D-89069 Ulm
- Building O27, 4th floor
- Room 424
Office HoursWith appointment.