Some questions and suggestions

[RoxaneSegers]

After some inspection of the owl conversion of the Conceptual Model made in Onto-UML, I have some questions and suggestions.:

-Apparently, multiple asserted class inheritance is allowed in Onto-UML. As such, there is also multiple asserted class inheritance in OWL. See for example: #Organisation which is modeled as subclass of #Agent and #FunctionalComplex. This implies that the class #Organization is shown twice when the model is loaded in Protege without any marker that these classes are equivalent. Is there a way to fix this, e.g. by preventing multiple class inheritance or by making the class equivalence explicit in OWL?

Modeling ’true’ individuals in the gufo subhierarchy of gufo#Individual. Is it possible to state in Onto-UML that some entity is an individual and not a concept? In the conversion, I see Named Entities such as ‘CAK’, ‘SVB’ which are names of specific organisations as classes. I would like to be able to say that these are individuals of #Organization.

-To the best of my understanding, Onto-UML is not geared towards a hierarchical modelling of concepts. As a result, the OWL conversion shows a rather flat hierarchy where I suppose that a deeper hierarchy is possible which has its benefits from a concise modelling perspective. See for example: #Inhabitant, #BSN holder, #Resident of The Netherlands, #Non-Resident, #Dutch Citizen which are sibling classes now.

-Is it possible to enforce in some way in Onto-UML that generic object properties such as #livesIn are defined with a domain and range at the highest possible level in the class hierarchy? In this case, the domain and range are (erroneously) defined as #ResidentInHealthcareInstitution and HealthcareProvider, while I suppose that this property generically applies between #Person and #Location.
Or, if one adheres to minimal ontological commitment, would it be possible to avoid domain and range and opt for a property restriction in the conversion of Onto-UML to gufo.owl?

-Is it possible to make e.g. symmetric properties explicit as such? See for example #partnerOf and #marriedTo.

[LucasBassetti]

Hey Roxane, thanks for your suggestion! Let me clarify a few points:

Apparently, multiple asserted class inheritance is allowed in Onto-UML. As such, there is also multiple asserted class inheritance in OWL. See for example: #Organisation which is modeled as subclass of #Agent and #FunctionalComplex. This implies that the class #Organization is shown twice when the model is loaded in Protege without any marker that these classes are equivalent. Is there a way to fix this, e.g. by preventing multiple class inheritance or by making the class equivalence explicit in OWL?

This is how Protegé represents the class with multiple inheritance and can be a little confusing indeed. You must look the URI. In this case, both Organizations are the same because they have the same URI.

Modeling ’true’ individuals in the gufo subhierarchy of gufo#Individual. Is it possible to state in Onto-UML that some entity is an individual and not a concept? In the conversion, I see Named Entities such as ‘CAK’, ‘SVB’ which are names of specific organisations as classes. I would like to be able to say that these are individuals of #Organization.

This is a OntoUML limitation. We will try find a solution for enable transform individuals in gufo.

To the best of my understanding, Onto-UML is not geared towards a hierarchical modelling of concepts. As a result, the OWL conversion shows a rather flat hierarchy where I suppose that a deeper hierarchy is possible which has its benefits from a concise modelling perspective. See for example: #Inhabitant, #BSN holder, #Resident of The Netherlands, #Non-Resident, #Dutch Citizen which are sibling classes now.

Those are modeling choices. You can change this by modeling the ontology in a more hierarchical way.

Is it possible to enforce in some way in Onto-UML that generic object properties such as #livesIn are defined with a domain and range at the highest possible level in the class hierarchy? In this case, the domain and range are (erroneously) defined as #ResidentInHealthcareInstitution and HealthcareProvider, while I suppose that this property generically applies between #Person and #Location. Or, if one adheres to minimal ontological commitment, would it be possible to avoid domain and range and opt for a property restriction in the conversion of Onto-UML to gufo.owl?

This is also a modeling choise. If this property is more generic you can change it in the original model to be between Person and Location.

Is it possible to make e.g. symmetric properties explicit as such? See for example #partnerOf and #marriedTo.

This is not possible in OntoUML at moment but we will try incorporate as a parameter in the transformation.