Open djogopatrao opened 9 years ago
pellet command line:
./pellet.sh query -q /tmp/query.sparql ~/git/tnm_ontology/ontologies/test.owl ~/git/tnm_ontology/ontologies/axioms.owl ~/git/tnm_ontology/ontologies/annotations.rdf ~/git/tnm_ontology/ontologies/tnm_6e_icdo_topographies.owl ~/git/tnm_ontology/ontologies/tnm_6th_edition.owl
We can test which test instances yield inference correctly by issuing a query such as below:
PREFIX rdfs: http://www.w3.org/2000/01/rdf-schema# PREFIX tnm: http://cipe.accamargo.org.br/ontologies/tnm_6th_edition.owl#
select ?pct (count(distinct ?ec) as ?ct) { ?ec rdfs:subClassOf tnm:EC. ?pct a ?ec. filter not exists { ?sub rdfs:subClassOf tnm:EC. ?pct a ?sub. ?sub rdfs:subClassOf ?ec filter ( ?sub != ?ec ) } } group by ?pct order by asc(?pct)
Instances with more than one class ?ec had the inference of EC class wrong. Instances with one instance had no inference at all.
It depends on resolution of #19