OnteEval Tools

OntEval is a ontology learning evaluation tool based on the Dellschaft and Staab's algorithm described in the [1] (ISWC 2006) On How to Perform a Gold Standard Evaluation of Ontology Learning. The tool takes two ontologys defined in either RDFS or OWL formats as input, one of which is assumed the Gold Standard ontology and the other as the machine computed ontology (or any other ontology you may wish to compare); then performs evaluation, computes measures including Lexical Precision, Lexical Recall, Taxonomic Precision, Taxonomic Recall, F-Measure and F'-Measure.

OntEval uses Jena Framework 2.5 to build tree structured ontology model from input ontology files. The tool supports RDFS and OWL formats, and for each format it supports a variety of model specifications as available in Jena OntModelSpec. The online tool uses OntModelSpec.OWL_MEM model specification for OWL and OntModelSpec.RDFS_MEM_RDFS_INF model specification for RDFS. If you wish to use different, or customised model specifications, or your input RDFS/OWL requires particular model specifications, please download the software and customise your own tool.

• Download software here; The package includes distributable jar's, dependency libraries, source code, javadoc, and sample testing files constructed using examples in [1] .

Perform evaluation online (temporarily offline)

[1] Dellschaft, K. & Staab, S. On How to Perform a Gold Standard Based Evaluation of Ontology Learning. International Semantic Web Conference, 2006, 228-241

This was developed as part of the Abraxas project, funded by the UK Engineering and Physical Sciences Research Council under grant number GR/T22902/0