Ontolearn is an open-source software library for explainable structured machine learning in Python. It learns OWL class expressions from positive and negative examples.
We need to implement subsumption based refinement operators (e.g. CELOE's) in different description logics (e.g. EL, ALC, SROIQ).
Ideally, we should be able to use these refinement operators across models without an effort.
By this, we could more objectivly quantify the benefits of choosen DL in learning OWL class expressions.
For instance, although EL is a subclass of ALC, to tackle a given learning problem (i.e., geneneralize well) we may not need to learn in ALC. Perhaps, EL would suffice. By this, we can compare the runtimes of using the same learner in different DLs.
We need to implement subsumption based refinement operators (e.g. CELOE's) in different description logics (e.g. EL, ALC, SROIQ). Ideally, we should be able to use these refinement operators across models without an effort.
By this, we could more objectivly quantify the benefits of choosen DL in learning OWL class expressions. For instance, although EL is a subclass of ALC, to tackle a given learning problem (i.e., geneneralize well) we may not need to learn in ALC. Perhaps, EL would suffice. By this, we can compare the runtimes of using the same learner in different DLs.