OM icon indicating copy to clipboard operation
OM copied to clipboard

Differentiation between cultural and physical meaning of length

Open jorisvanduijneveldt-tno opened this issue 1 year ago • 5 comments

The ontology provides an om:Length class which is a purely physical concept (the amount of space between two points along a curve). It further provides the subclasses om:Width and om:Height which provide additional meaning to the type of length. I’m missing a way to capture the cultural/conceptual meaning of length associated with specific objects.

This poses difficulties in two scenarios:

  1. Enforcing three-dimensional presence: Ensuring objects like cars always have three defined dimensions (length, width, and height) is currently challenging.

  2. Specific queries: Performing queries specifically for the "length" of an object (excluding width or height) becomes ambiguous.

I'm seeking suggestions on how to address this issue and potentially identify a solution or workaround to represent the cultural/conceptual meaning of length within the ontology.

jorisvanduijneveldt-tno avatar Feb 15 '24 11:02 jorisvanduijneveldt-tno

Very interesting subject, thanx for raising this. I see the relevance. Would you advocate not working with subclasses of quantities? In principle, the dimension could (or would?) be a (the?) way to relate e.g. length, width, height, etc. I think this is something we have to consider in the light of the future development of OM 3.0.

HajoRijgersberg avatar Mar 10 '24 15:03 HajoRijgersberg

I think there definitely is value in using subclasses for quantities. In our project we defined a new length as subclass of om:Length in a separate namespace, making it a sibling of width and height. Ideally I would love such a subclass in OM 2.0 but I don't have a good suggestion for the identifier.

Using dimensions to relate different quantities would indeed group length, width, and height. I don't see this a being the solution without consequences. Quantities can share dimensions but have a very different meaning. A few examples would be units including angles or solid angles (which have a unit dimension), om:Torque and om:Energy, om:EnergyDensity and om:Pressure.

jorisvanduijneveldt-tno avatar Apr 08 '24 08:04 jorisvanduijneveldt-tno

How is that subclass called exactly?

Indeed, I also see the danger with dimensions. Thinking wildly,, would it be an idea if we defined specific kinds of volumes as multiplications of e.g. length, width and height, but also e.g. length, width and depth, and others probably?

HajoRijgersberg avatar May 12 '24 08:05 HajoRijgersberg

Using our own namespace, we have the luxury to reuse the name :Length which of course is not an option within OM. An alternative could be om:LongitudinalLength. I think this would correctly convey that it describes an om:Length in a specific direction.

jorisvanduijneveldt-tno avatar May 24 '24 09:05 jorisvanduijneveldt-tno

So, in your namespace you can classify like (where the bullets indicate subclasses and their depth):

Length

  • Length
    • Length of people
      • length of Joris
  • Depth
    • Depth of people
      • depth of Joris
  • etc.

In OM we can only do like the following, which (indeed) I see as a disadvantage because 'Length of people' and 'Depth' are at the same level:

Length

  • Length of people
    • length of Joris
  • Depth
    • Depth of people
      • depth of Joris
  • etc.

I would like to have your feedback wether I see the above correctly (i.e., in your namespace), and if you think that how it is in OM is indeed a disadvantage. So, if I understand you correctly you indicate that Length could perhaps be called LongitudinalLength. My first feeling says that that is more like a geo term. But I'll think about it. Thanx for the suggestion.

HajoRijgersberg avatar Jun 22 '24 09:06 HajoRijgersberg