`madeExecution` `madeBySystem` super-properties used on `Execution`

[dr-shorthair]

There is an asymmetry between

• usedProcedure, that links any instance of an Execution or sub-class to its Procedure,

and

• madeByActuator madeBySensor madeBySampler that links an instance of a specific Execution sub-class (i.e. Actuation Observation Sampling) to an instance of the System subclass (i.e. Actuator Sensor Sampler) that made it.

Should this be neatened up by introducing madeBySystem as the super-property? And usedActuatingProcedure usedObservingProcedure usedSamplingProcedure as sub-properties?

[maximelefrancois86]

+1 on this. Anything that improves homogeneity is desirable. I expect there may be some objections to using subproperty axioms in SOSA, though.

[dr-shorthair]

As per precedent, the sub-property axioms would be in ssn.ttl, and not in sosa.ttl

[dr-shorthair]

madeBySystem and madeExecution are logical requirements to accompany https://github.com/w3c/sdw-sosa-ssn/pull/130

[dr-shorthair]

Terms in red are the proposed additions, classes in brown are changes we already introduced

[dr-shorthair]

@alexrobin how do these proposals line up with Connected Systems?

[alexrobin]

Yes, but in fact we only use the equivalent of the super-properties madeBySystem and usedProcedure since the surrounding object provides the context.

[dr-shorthair]

Specific predicates madeByActuator/madeActuation , madeBySensor/madeObservation , madeBySampler/madeSampling must be retained for backward compatibility reasons.

With the introduction of Execution I suggest that the pair madeBySystem/madeExecution should be added. These fill in some gaps in the overview diagram for Execution, as shown.

But I tend to agree that usedActuatingProcedure , usedObservingProcedure , usedSamplingProcedure may not be needed, and risk bloating the ontology. Of course the range of usedProcedure can be constrained by Restrictions on the specializations of Execution (i.e. Actuation, Observation , Sampling) in the 'SSN' graphs.

[dr-shorthair]

• We already agreed to add the classes with Brown names.

• I suggest we add the Red predicates.

• While they would make the ontology more homogeneous overall, I'm now of the opinion that the Pink predicates are not needed and would unnecessarily bloat the ontology.

[dr-shorthair]

@ldesousa seems to generally agree to limit the additional terms, in his comment on the draft PR

[dr-shorthair]

Preview here: https://raw.githack.com/w3c/sdw-sosa-ssn/225-madebysystem/ssn/index.html#SOSAmadeBySystem