Ontology reasoning

[naparnanayak]

I have extented the DMOP ontology for meta-learning using protege. I am able to run the reasoner on TBox.

I am populating this ontology using owlready. However, when I run the reasoner I get the following error:

Hermit:



Pellet:



There is nothing on the specified line.

I do not have any rules defined in the ontology at this point.

Any idea how can I solve? Because I have multiple datapoints, I am unable to paste the entire code here.  But here is the snippet.

Assume I have loaded all the namespaces. I am able to save this generated RDF file, but not the reasoning.

When I run Hermit or Pellet using protege on the knowledge base the protege classifies classes and object properties. While running the reasoner on data properties, it gets stuck.

for i in range(len(df)):
        featureSelectionTask = dmop.FeatureSelectionTask()
        mtla = mtl.MetaLearningAlgorithm()
        fileName = df['File'][i]
        fileObject = dmop.DataSetClass(fileName)
        mtla.hasMetaObjective.append(featureSelectionTask)

        featureSelectionTask.specifiesInputClass.append(fileObject)
       
        outClass =dmop.StructuredPredictionModelClass()
        featureSelectionTask.specifiesOutputClass.append(outClass)
        outClass.hasValue.append(df['FeatureAlgo'][i])
       
        nFeatures = dmop1.NumberOfFeatures()
        fileObject.hasQuality.append(nFeatures)
        nFeatures.hasValue.append(int(df['nr_attr'][i]))
        nFeatures.hasBinValue.append(str(df['nr_attr_bins'][i]))
        ......



I solved the pellet issue by setting JAVA_HOME variable for owlready. Still I have the same problem for Hermit.

[Jiba]

Hi,

For solving the Hermit Problem, you need to obtain the full Hermit error message.

This can be done by adding the "keep_tmp_file = True" argument when calling sync_reasoner(), and then calling the reasoner manually in the terminal (you can copy paste the command line shown in the Owlready messages).

Jiba