Property request: has immaterial entity host

[alanruttenberg]

Relates a site or boundary to the entity relative to which it is defined

I use this, for example, to relate something like the Greenwich meridian or WGS84 Ellipsoid (boundaries) to Earth.

[alanruttenberg]

domain: site or continuant fiat boundary
range: material entity

[neilotte]

@alanruttenberg Can we try to refine this definition? e.g. what does it mean to be 'defined' in this sense?

'Entity' also seems too broad (e.g. that would include spatial regions). There's already relationships between processes and where they occur. Can I recommend this be refined?

[APCox]

I'm finding the relation name confusing. "has immaterial entity host" suggests to me that the host is an immaterial entity, but in your example the host is Earth, which is a material entity. Propose changing to "immaterial entity has host" or simply "has host" (though the latter will no doubt cause a different confusion). The inverse would be "host of immaterial entity".

[alanruttenberg]

Not averse to renaming in principle. "has host" is too broad I think, although maybe as altLabel. In general it's a good idea to have the relation fit into a sentence. Compare:

1. site of my room at home has immaterial entity host my home.
2. site of my room at home immaterial entity has host my home.

(2) not so good. Maybe "hosts immaterial entity" for the inverse, which I didn't specify. Or "is host of immaterial entity" using some of Alex's language. I think (1) is still better. Another might be "has host of immaterial entity"

site of my room at home has host of immaterial entity my home

To the definition, here are the definitions for site and boundary:

Site

b is a site means: b is a three-dimensional immaterial entity whose boundaries either (1) (partially or wholly) coincide with the boundaries of one or more material entities or (2) have locations determined in relation to some material entity [034-BFO]

Continuant Fiat Boundary

b is a continuant fiat boundary means: b is an immaterial entity that is of zero, one or two dimensions, which is such that there is no time t when b has a spatial region as continuant part at t, and whose location is determined in relation to some material entity

Revised definition, reusing the words from the definitions:

Relates a site or boundary to the material entity that its location is determined in relation to, or a site to the one or more material entities whose boundaries wholly or partially coincide with its boundary.

[dlutz2]

Two questions:
Do we wish to distinguish the two cases "coincide with boundary" and "defined in relation to" as distinct properties? The first seems to be very specific topological relation(s) while the other could be anything.

Speaking of which, is "defined in relation to" restricted to a spatial/topological relations? For example:

1. 300 meters above the earth
2. a void within the material portion of an material entity
3. the volume visible from the driver's seat
4. all the (sites of) real estate own by an organization
5. all areas contaminated by a toxin

1-2 seems strictly geometric/topological, 3 is a combination of domain and geometric, 4-5 are completely domain.
Having a distinguished place to hang domain-specific relations like "visible from", "contaminated by", "owned by", ... would be useful.

[alanruttenberg]

Two questions: Do we wish to distinguish the two cases "coincide with boundary" and "defined in relation to" as distinct properties? The first seems to be very specific topological relation(s) while the other could be anything.

Answers are off the top of my head, intended to spur further discussion rather than being definitive.

I have no need to distinguish, but we can postpone deciding until there is an affirmative requirement, in which case we could add two subproperties of the proposed relation under the interpretation that "defined in relation to" is not your "domain" but essentially records that there is a relational position quality that specifically depends on the site and the material entity.

I don't need to distinguish between those because my usage is just to make clear what the site or boundary is tethered to.

Speaking of which, is "defined in relation to" restricted to a spatial/topological relations? For example:

Short answer: I think yes. TL;DR:

1. 300 meters above the earth

This seems clear enough and is, I think, a prototypical case.

2. a void within the material portion of an material entity

The relation would be to the material entity. Again seems prototypical

3. the volume visible from the driver's seat

This one is harder. I think I would consider one host to be Earth, and the other the eyes of the passengers.

4. all the (sites of) real estate own by an organization

It isn't clear whether a site is intended to be contiguous but my inclination would be it is. All the sites would then be some sum, a subclass of immaterial entity but not any of its current subclasses. I think both the individual and sum would have the same host: Earth. Or some part if you want to be specific.

5. all areas contaminated by a toxin

Similar to 4 in that it is a sum. The individuals would have material hosts that I think would be the range of material contaminated by the toxin. The sum is harder. Maybe a material entity that is the sum of the others, but again not a subclass of any of the current material entity subclasses. Fiat sum of fiat parts.

1-2 seems strictly geometric/topological, 3 is a combination of domain and geometric, 4-5 are completely domain.
The cases where you are defining a an entity as a sum of things for which a predicate is true is called a fusion, I believe. I believe I know how to express that in FOL. In OWL I think you would represent it with a relation to the relation(predicate)
Having a distinguished place to hang domain-specific relations like "visible from", "contaminated by", "owned by", ... would be useful.
'Visible from' is a use case in my domain. I currently define sensor field view as 'A site, whose surrounding medium, what's located in it, and events which occur in it are detectable by a sensor' I don't currently have a relation to the sensor, but thinking about it, I don't think it would be wrong to say one of it's host is the sensor. On that interpretation 'is visible from' is a subproperty of 'has immaterial entity host' restricted to sensors.
If it images the ground in part, then another host would be a Portion of Geosphere.

In the case of ownership there would be a mutually dependent role between owner and owned. So the path would be site (or what site is part of) -> owned role -> owner role -> owner.

Contaminated by the path is site -> host -> contaminated quality (relational) -> other bearer (whatever is contaminating)

Seems like the desired parent relation would be a shortcut for these. We should probably copy this into a new issue since it is more about defining sites or what they are related to than about the relation of site to host.

On the initial question, whether these all fall under "defined in relation to", I don't think it isn't the intention that "in relation to" it means stuff like owner or contaminant - your domain elements. Your domain-based definitions seem more in the realm of defining sums as { x | p (x) }, which suggests to me that we should be able to define an entity with some relationship to p.