Abstraction of the Location_Of Meronymy

[jeshaw2]

What would you consider to be a good abstraction of the Location_of Meronymy?   This is a request for comments.

This is another thread in my concers about abstractions of the six types of meronymy.  The first thread of this series is the Part_of Meronymy .  The introductory and Background portions of that thread should be reviewed by readers new to this series.

Now to the Abstraction...A work in progress.

Precise place (Location) / area
Locations do not really contribute to an area in a functional way. This sub relation expresses spatiality and temporality, as in : oasis/desert, Alps/Europe, month/day/year.

• Area boundary coordinates are dynamic., [Policy}
• Areas form a system of overlays based on their granularity.
• Areas may partially, or completely, overlap each other.
• Areas may participate as locations in a larger area.
• Locations may span more than one area, as in mountains ranges within states.[Policy]
• Areas and locations may have several aliases. [Policy]
• There is always an area at the most granular levels of the spatial or temporal hierarchy that fully contains a given location.
• Destruction of an area destroys its associated locations unless those locations have a mapping into another area.
• Destruction of all areas into which a location may be mapped  also destroys the location.
• Locations need to know what areas they are in.
• Areas need to know what locations are within their boundaries.
• Locations may be in several areas at the same time.
• Mapping rules allow conversion of a location’s coordinates in one area into the coordinate system of an overlapping area.
• Areas do not own locations within their boarders.

Notation: An Association stereotyped as <<locationOf>>.  Association ends may be adorned with navigation arrowheads and ownership diamonds.  A link name (formed by the conjunction of two verbs, or verb phrases, separated by a slash) assists the reader in verbalizing the association from either of its ends. Serious consideration should be given to the use of an Association Class to provide information about, and methods to manage, the nature of the association.

[Paolo F Cantoni]

Jim,

I think this one needs a make-over.  I think this misses the point about «locationOf».

Firstly, my references don't include temporality, only spatiality.
The reason, I think, is related to the view that «locationOf» is probably the weakest of the meronymic relationships.  In other words, if you can find a stronger relationship between the whole and the part then you should use that one.  For example, if you take the 27th out of February - you no longer have February...  Thus, you should use «partOf» in preference.

HTH,
Paolo

[jeshaw2]

I heartily agree that this needs  work.  I'm posting these threads to generate discussion and by them gain much new knowledge.

Allow me to share my view point and perhaps inject a bit of levity into this heavy stuff.  

I think this misses the point about «locationOf».

And the point is...?

The reason, I think, is related to the view that «locationOf» is probably the weakest of the meronymic relationships.

Respectfully, Who's view?  I'm not sure that MapQuest, Human Resource departments, Geneologists, Theoretical Physicists, and Law Enforcment Officials would agree that the association is a weak catch-all relationship.  I've used it often while modeling a logging of an individual's residence history and in modeling complex sales territories for Sales Commission applications.  This type of relationship is pivotal in the design of these applications.

Firstly, my references don't include temporality, only spatiality.

Yes, my books were like that too, but then I bought a book by Stephan Hawking.  You know, the Space-Time Continuum guy?   He published after the Meronymy papers were published, so (in the spirit of Henry Leonard and Nelson Goodman) I updated the meronymy concept to make it as current and abstract as I could make it.  

For example, if you take the 27th out of February - you no longer have February

Exactly!  In my calendar, you can take out the 29th (mutating the leap-year into a non-leap-year [Policy]) and still have a February, but If I remove the 27th, I no longer have a February, just a range of days (as in the Julian calendar).  My abstraction rules allow this. Now if you are suggesting that in your calendar, after removing the 27th you still do have a February, then your calendar defines a different area of the space-time continuum than mine and we need to discuss some mapping rules.  

Thus, you should use «partOf» in preference.

Interesting... Are you leading me into a discussion of the Composition of the Space-Time Contiuum?  I've got books on String Theory too.  He he, just joking!

I apologise for pulling your leg(s).   I am here to learn. Can you set me on an improved path?

[Paolo F Cantoni]

Hi Jim,

Just a quick note.

I think we need to make sure we aren't confusing the "X" must be located in "Y"  (Association name and semantics) with the «locationOf» Association Type?

If there is a functional relationship then «locationOf» is inappropriate.  In my view, it's there to indicate that the only reason for the relationship is that the part "just happens to be in the whole (by location)".  Hence the example of oasis/desert.  That's why I viewed it as the weakest...

Where more formal relationships exist (such as between Local Government Area (City, US County, Shire etc), National Administrative Region (State, Province, UK County etc), and Country, then «locationOf» is inappropriate.  But I'm prepared to be converted...

Paolo

[jeshaw2]

Paolo;

I'm not sure I see a difference in the two things you are talking about...

I think we need to make sure we aren't confusing the "X" must be located in "Y"  (Association name and semantics) with the «locationOf» Association Type?

Is must the operative word here?  My vision does not include any such rules.

Where more formal relationships exist (such as between Local Government Area (City, US County, Shire etc), National Administrative Region (State, Province, UK County etc), and Country, then «locationOf» is inappropriate.

Can you expand on the concept of a formal relationship?  Just as there are legal documents that place a City in a State, there are legal documents that place that same City in a Geographic area.  If you are talking about the structure of government bodies (City government within a State Government, and other metonymic uses of a location's name; (e.g.; using a location's name to represent the governing agency of the location)), I agree that other associations are more appropriate.  I also agree if the sub-relation is attempting to describe forms of collaboration or how areas are dependent upon each other in some way.  

In my conceptualization, the only functionality or behaviors relevant to this sub-relation would be held in an AssociationClass; their primary function being the mapping of locations from one area to another and/or discovering the border coordinates of a common area resulting from the union of the two other areas.

Do you feel a need for conversion?  

[sargasso]

While <locationOf> might be semanticly weak, only offering a percentage increase over <partOf> it can often provide strong value in constraint determination.

Iff <locOf> LastCanOfBeer is InTheFridge then me.consume(beer) can operate effectively and with value.  OTOH if <locOf> YourCheque is InTheMail then there is precious little operations of value that I can apply.

bruce

[Paolo F Cantoni]

While <locationOf> might be semantically weak, only offering a percentage increase over <partOf> it can often provide strong value in constraint determination.

Iff <locOf> LastCanOfBeer is InTheFridge then me.consume(beer) can operate effectively and with value.  OTOH if <locOf> YourCheque is InTheMail then there is precious little operations of value that I can apply.

bruce

bruce,
your response once again raises the question of is located vs «locationOf».  For an Association to be a «locationOf» meronymy, there has to be a whole for which there is a part.

Thus the relationship of can to slab (while it is still in the slab - a short period of time where you're concerned, I know  )  is a meronymy.  The relationship of can to fridge is not.

Where something is located (in the absence of a meronymy) is just a normal Association.

Paolo

[sargasso]

Now you've confused me.  

[Paolo F Cantoni]

Now you've confused me.  

In this thread, Jim is discussing the Location_Of meronymy (which in line with our - Jim and my - discussions) we have agreed to attribute the «locationOf» stereotype.

Meronymy is the relationship between the part and the whole.  The meronym is the part, the holonym is the whole.  Therefore, in the case of the «locationOf» meronymy, both have to be locations.

Thus the oasis is the part in the desert.

When we want to know where something is located, then we can simply use a normal association, naming it accordingly.

The confusion arises in that just about every «locationOf» meronymy can be named is located... but not every is located... is a meronymy.

Now, it is my assertion, that where you have a «locationOf» meronymy, you often have a stronger meronymic relation which should take precedence.  That's why I assert it's a weak association, and basically should only be used where no other meronym is present.  This particular oasis just happens to be located in this particular desert.  But oases are always in deserts (hence the meronymy - desert: holonym, oasis meronym).

HTH,
Paolo

[jeshaw2]

Thus the relationship of can to slab (while it is still in the slab - a short period of time where you're concerned, I know  )  is a meronymy.  The relationship of can to fridge is not.

I'm nor sure what a slab is, but I'm guessing it is some type of can or bottle?  Over here, a slab is a table located in the morgue where cadavers are placed.  Perhaps slag is a symbolic of drinking excessive beer? Anyhow, based upon the assumption that a slab is a can:

Are you sure the relationship of can to slab is a meronymy and not beer in a container?  The beer is not intrinsic to the can, but then, when I purchase the can, I do not consider the beer to be an adornment either.  After I drink the beer from the can, the can is crushed (destroyed), not by my removing the beer, but by the reaction to my frustration at having no more beer to drink.

Meronymy is the relationship between the part and the whole.  The meronym is the part, the holonym is the whole.  Therefore, in the case of the «locationOf» meronymy, both have to be Spatial or Temporal locations.

The text in Blue are my revisions to the original statement.

This particular oasis just happens to be located in this particular desert.  But oases are always in deserts (hence the meronymy - desert: holonym, oasis meronym).

Paolo, can you expand on this?  Is always definitive of the meronymy?  I guess what you are saying is that the meronymic relationship must be true at the class level, not just at the level of specific objects.

Now, it is my assertion, that where you have a «locationOf» (Spatial or Temporal) meronymy, you often have a stronger meronymic relation which should take precedence.  That's why I assert it's a weak association, and basically should only be used where no other Spatial or Temporal meronym is present.

An example of this might help me to better understand your assertion.    Spatial & Temporal are definitive to this meronymy, so what other meronymy type could there be that is stronger?

[KP]

I'm nor sure what a slab is, but I'm guessing it is some type of can or bottle?

No, a slab is the outer packaging, typically 24 cans shrink-wrapped together.

Well actually that's not quite right... it's not the outer packaging, because without the cans it isn't a slab it's just a piece of tatty cardboard and some polythene.

I don't know how this affects the points you were making?

[Eve]

The slab (in my opinion) is the collection of the cans.  To the extent that if you remove a can or two, I'm not sure if you have a slab left.  I don't think you'd get away with bringing a slab minus some cans if you had previously said you'd bring a slab.

Now, I guess a can can have a location_of relationship to a slab, but it seems to me that part_of or portion_of would be a more appropriate relationship to use.  Not that I've considered all of the relationships in full detail.

Disclaimer.  These comments made by someone with very limited knowledge of the problem domain.

[Paolo F Cantoni]

The slab (in my opinion) is the collection of the cans.  To the extent that if you remove a can or two, I'm not sure if you have a slab left.  I don't think you'd get away with bringing a slab minus some cans if you had previously said you'd bring a slab.

Now, I guess a can can have a location_of relationship to a slab, but it seems to me that part_of or portion_of would be a more appropriate relationship to use.  Not that I've considered all of the relationships in full detail.

Disclaimer.  These comments made by someone with very limited knowledge of the problem domain.

Excellent Simon!

I was waiting to see if any one else "got it"!

BTW: my bet would be on Portion_Of...  Since you can vary the number of cans in the slab.  Part_Of (as we've previously discussed) means if you take one part, you no longer have the whole.  Since I can have a 24 can slab or a 30 can slab, that doesn't seem to apply.

Paolo

[jeshaw2]

IMHO all of the meronymic relationships available to us here.  The ones to use, of course, depend upon the modelers focus.

If your focus (as a beer consumer) is upon just the beer and not the containers, then the <<portionOf>> relation is appropriate; portions of the total volume of beer in the slab.  That the portions are individually wrapped is immaterial here.

If your focus (as a part of the distribution chain) is on the slab inventory, the <<partOf>> relation is appropriate.  A cardinality of [24, 30] would solve the issue of the number of cans in the slab, but it would be better to specialize slab as 24-canSlab and 30-canSlab for pricing and inventory control reasons.  Either way, removing the tattered cardboard & plastic packaging material would destroy the slab yielding individual cans.

If you have been drinking beer for many years and are concerned about the ever growing "Love Handles" about your waist, then the <<substanceOf>> relationship is important.  I'm assuming here that the ingredients list and nutritional information is available either on the slap packaging or on the individual cans.

If you have come lately of age to drink beer, you may be focused on the "manly" way to drink beer (can opening, guzzling vs straw sipping, management of bodily functions, singing & dancing, can disposal, etc.) then the <<subActivityOf>> is appropriate.

If your focus is on the social gathering for which you are supplying the slabs, then <<memberOf>> is appropriate.  Member's age, beer preference (light or dark), and capacity may be of interest relative to the properties of the slabs being provided.

If the purpose of the social gathering is the enjoyment of a Kabuki Theater production (an event typically lasting 7 or more days) and beer drinking is confined to periods of intermission,  then <<locationOF>> is appropriate.  Here we are concerned about the temporal location of the intermission periods within the total temporal period of the production.  We would then map those temporal locations in to the temporal periods of our distillery & logistics calendar. That way, the beer may arrive fully chilled and at the start of each intermission period.

So, if you guessed any of these, you are correct!  
Congratulations!  

[sargasso]

  All of which just goes to show how important beer is!