skos <> ontology combi

[Bohms, H.M. (Michel)]

Is it save to combine skos and rdfs/owl?

 

Ie have an rdfs/owl ontology directly import an skos vocabulair.

 

I see examples:

https://www.w3.org/TR/2009/REC-skos-reference-20090818/#L896,

 

I can remember we earlier said: better to keep really separate and link via say rdfs:seeAlso (from class to concept) but it would be convenient to just have:

 

<MyConcept> rdf:type skos:Concept , owl:Class .

 

Thx for your view, Michel

 

Michel Böhms Data Scientist Unit Buildings, Infrastructure & Maritime Mobile +31 630381220 E-mail  michel...@tno.nl Location   This message may contain information that is not intended for you. If you are not the addressee or if this message was sent to you by mistake, you are requested to inform the sender and delete the message. TNO accepts no liability for the content of this e-mail, for the manner in which you use it and for damage of any kind resulting from the risks inherent to the electronic transmission of messages.

 

 

[David Price]

SKOS is itself is just an OWL ontology. It is OWL Full though.

However, A thing  being a Class and an instance of Concept can be a problem in some scenarios. For example, it means both subClassOf and broader/narrower are both usable for that thing. What does that even mean?

FWIW a common case of using SKOS and another ontology together is when a SKOS-based taxonomy is used to manage categories, for example, that are then used as a drop down list for a property value. In that case subclasses of Concept are used to group the categories and provide a class for the property value range. A second common case is when the taxonomy concepts are used to tag things as a search or navigation aid.

Cheers,

David

On 3 Jan 2023, at 09:41, 'Bohms, H.M. (Michel)' via TopBraid Suite Users <topbrai...@googlegroups.com> wrote:



Is it save to combine skos and rdfs/owl?

 

Ie have an rdfs/owl ontology directly import an skos vocabulair.

 

I see examples:

https://www.w3.org/TR/2009/REC-skos-reference-20090818/#L896,

 

I can remember we earlier said: better to keep really separate and link via say rdfs:seeAlso (from class to concept) but it would be convenient to just have:

 

<MyConcept> rdf:type skos:Concept , owl:Class .

 

Thx for your view, Michel

 

Michel Böhms Data Scientist Unit Buildings, Infrastructure & Maritime Mobile +31 630381220 E-mail  michel...@tno.nl Location

<image001.gif>

This message may contain information that is not intended for you. If you are not the addressee or if this message was sent to you by mistake, you are requested to inform the sender and delete the message. TNO accepts no liability for the content of this e-mail, for the manner in which you use it and for damage of any kind resulting from the risks inherent to the electronic transmission of messages.

<image002.png>

 

 

--
You received this message because you are subscribed to the Google Groups "TopBraid Suite Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to topbraid-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/topbraid-users/eadf07387dc54e35b4cc5ad963cddf5c%40tno.nl.

[Bohms, H.M. (Michel)]

The example below shows perfectly ok in TBC.

Ok in your view?

 

 

# baseURI: https://w3id.org/sml-example/skos-rdfs/def

# prefix: ex

 

@prefix ex: https://w3id.org/sml-example/def# .

@prefix owl: http://www.w3.org/2002/07/owl# .

@prefix rdf: http://www.w3.org/1999/02/22-rdf-syntax-ns# .

@prefix rdfs: http://www.w3.org/2000/01/rdf-schema# .

@prefix skos: http://www.w3.org/2004/02/skos/core# .

@prefix xsd: http://www.w3.org/2001/XMLSchema# .

 

https://w3id.org/sml-example/skos-rdfs/def a skos:ConceptScheme, owl:Ontology .

 

ex:Bridge

  a rdfs:Class, skos:Concept ;

  skos:prefLabel "Bridge"@en ;

.

ex:LoadLevelType

  a rdfs:Class, skos:Concept ;

  skos:prefLabel "Load level type"@en ;

.

ex:Heavy

  a ex:LoadLevelType, skos:Concept ;

  skos:prefLabel "Heavy"@en ;

.

ex:Normal

  a ex:LoadLevelType, skos:Concept ;

  skos:prefLabel "Normal"@en ;

.

ex:Light

  a ex:LoadLevelType, skos:Concept ;

  skos:prefLabel "Light"@en ;

.

ex:loadLevelType

  a rdf:Property ;

  rdfs:range ex:LoadLevelType, skos:Concept ;

  skos:prefLabel "load level type"@en ;

.

ex:MyBridge

  a ex:Bridge ;

  ex:loadLevelType ex:Normal ;

.

 

Michel Böhms Data Scientist Unit Buildings, Infrastructure & Maritime Mobile +31 630381220 E-mail  michel...@tno.nl Location

This message may contain information that is not intended for you. If you are not the addressee or if this message was sent to you by mistake, you are requested to inform the sender and delete the message. TNO accepts no liability for the content of this e-mail, for the manner in which you use it and for damage of any kind resulting from the risks inherent to the electronic transmission of messages.

 

To view this discussion on the web visit https://groups.google.com/d/msgid/topbraid-users/CBDF1E23-8F94-4F72-BAA4-3181326BE22C%40topquadrant.com.

[David Price]

On 3 Jan 2023, at 13:42, 'Bohms, H.M. (Michel)' via TopBraid Suite Users <topbrai...@googlegroups.com> wrote:

The example below shows perfectly ok in TBC.

Ok in your view?

IMO No. Others may disagree, of course.

TBC is very flexible as is RDF itself as far as what triples it will let you create. As I mentioned, my issue is about the semantics of these statements. IMO mixing things as being Class AND SKOS Concept is bad modelling practice. When do you use subClassOf vs broader/narrower .. or even rdf:type vs broader/narrower.  Will the tool into which you might feed this RDF complain about individuals also being classes when it is clearly NOT a case of punning. Tools like EDG, for example, expect a separation between "data model" and “data" in order to generate UI frameworks (GraphQL schemas for EDG) and for performance optimisations … which this violates. For EDG Data Graphs to work you’d also need to make the classes SHACL NodeShapes so it can generate the framework for its model-driven UI.

  

FW|IW RDF being so flexible is a “problem” at times. RDF does not understand the semantics of languages built on top of it which leaves following best practices up to the modellers. Just because a tool does not stop modellers from doing less-than-ideal things does not mean they should do those things. 

Again … just my opinion.

FYI If one were following the scenario 2 I mentioned, LoadLevelType would be a subclass of Concept, not a member of Concept.  I guess the obvious question is … if you are not using any of the SKOS semantic properties, then why import and embed the whole of SKOS in this picture? If you just want prefLabel, etc annotations then just add that property to your ontology (i.e. copy and paste from SKOS and adapt for your case).

Cheers,

David

To view this discussion on the web visit https://groups.google.com/d/msgid/topbraid-users/45d3acce580f467d953cefba946e5c3b%40tno.nl.

[Bohms, H.M. (Michel)]

Thx David for your views.

 

For now we stick to separation of worlds (like in our CEN TC442 SML standard).

If this is giving issues we best follow your advice in your last sentence.

(forget about skos concepts, just use the annotation skos props in the ontology…).

 

I can remember earlier discussions with Jan Voskuil of Taxonic that came to same conclusion if I remember well.

(in the end your are mixing two worlds, the words of things and the world of terms for things that might be rdf-technically ok but conceptually flawed)

 

Thx Michel

 

Michel Böhms Data Scientist Unit Buildings, Infrastructure & Maritime Mobile +31 630381220 E-mail  michel...@tno.nl Location

This message may contain information that is not intended for you. If you are not the addressee or if this message was sent to you by mistake, you are requested to inform the sender and delete the message. TNO accepts no liability for the content of this e-mail, for the manner in which you use it and for damage of any kind resulting from the risks inherent to the electronic transmission of messages.

 

To view this discussion on the web visit https://groups.google.com/d/msgid/topbraid-users/DC039B00-661F-4DD9-ABA8-EE9AB112CA5D%40topquadrant.com.