[Top] [All Lists]

Re: [ontolog-forum] N-RELATIONs: Formal Ontology, Semantic Web and Smart

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: "doug foxvog" <doug@xxxxxxxxxx>
Date: Mon, 7 Nov 2011 16:51:27 -0500 (EST)
Message-id: <61155.>
On Mon, November 7, 2011 14:38, AzamatAbdoullaev said:
> "Expressing the verb phrase "x isMarriedTo y" or "x hasSpouse y" or
> "Marriage(x,y) are saying the SAME THING, it is the same fact
> expressed in a different way - in some cases as relations
> and some cases reified.    (01)

I agree.    (02)

> I submit
> that to understand how the different ways we express information are
> related we have to understand concepts at this level."    (03)

If you mean at the reified level, INSTEAD OF at the relation level,
i agree.    (04)

> That's the job ontology to identify the nature of concepts or the real
> semanics of terms, i.e., what real things they signify, denote and
> connote; or mean, represent and sense.
> Indeed, "marriage" denotes a relationship, a spousal relationship,    (05)

It is important to distinguish between a spousal relationship and
an ontological one.    (06)

> as its primary meaning, its barest definition, with
> all its possible types: as common-law marriage, marriage of convenience,
> plygamy.    (07)

As such, it is a full-fledged ontological object with many possible
(ontological) relations between it and other things.    (08)

> But it    (09)

By "it" do you mean the WORD "marriage" or the CONCEPT of marriage?    (010)

> connotes all other things associated with it, married
> couple, man and wife, wedding or marriage ceremony, and many other
> associations.    (011)

I would hold that these are all features of the CONCEPT of marriage,
and as such are connoted by any WORD which denotes or connotes marriage.    (012)

> At the level of language, natural or artificial, the term/concept    (013)

The TERM and the CONCEPT are two very different things.  One is a
lexical concept; while the other is a type of situation.    (014)

> could be virtually expressed in infinite ways. But still,
> it's a binary relation.    (015)

I assume by "relation", you mean a kinship relationship between two
people.  One thing that can be stated about such a kinship relationship
can be expressed by an ontological binary relation which could be
labeled "spouseOf" or some such.  But the relationship is a far richer
conceptual object than merely a single fact that can be expressed as
a binary relation holding between two people.    (016)

[Comments continued below.]    (017)

> ----- Original Message -----
> From: "Cory Casanave" <cory-c@xxxxxxxxxxxxxxx>
> Sent: Saturday, November 05, 2011 9:38 PM    (018)

>> Doug,
>> Whenever I see properties with complex verb phrases like "
>> isCurrentlyMarriedTo ", this is a red flag that indicates that multiple
>> concepts are being conflated.    (019)

Which is a reason that i criticized such a predicate as the sort of
thing created by "beginning ontologists" who don't "stop[] to consider
what the underlying classes of things are [or] realiz[e] that many
more relations could apply"    (020)

>> Lets say this was expressing X isCurrentlyMarriedTo Y.
>> There are multiple possible facts and representations here:    (021)

You providing multiple representations of the same fact below
as well as representations of related facts.    (022)

>> * X is married
>> * Y is married    (023)

These two facts can be concluded from the below fact.    (024)

>> * The spouse of X is Y
>> * X hasSpouse Y
>> * X is married to Y    (025)

The above three sentences have the same meaning.    (026)

>> * The above situation currently exists    (027)

No "situation" has been defined.  That is a problem which i discussed.    (028)

>> * The spouse of Y is X    (029)

If this relation was defined as symmetric, than this is another redundant
assertion.  If not, the ontology is faulty.    (030)

>> * There is a Marriage, in which X and Y are spouses    (031)

This is the model that i was suggesting.  Note that using this model,
you are making three assertions.    (032)

>> * M exists now.    (033)

A fourth assertion, although the contextual Now needs to be defined.    (034)

>> * X hasSpouse Y    (035)

You already stated this    (036)

>> * Y hasSpouse X    (037)

Another rewording.    (038)

>> Perhaps X is the male in a traditional marriage.    (039)

>> * X isHusbandOf Y
>> * Y isWifeOf X    (040)

This is something else missing from the criticized predicate.    (041)

>> Etc, Etc.    (042)

>> ...
>> The problem with this is that these made-up verb phrases are black boxes
>> that become difficult to relate to the other made-up verb phrases.  I
>> suggest we have to stop inventing new "words" and make more of the
>> concepts we have.    (043)

It is not "words" that are needed, but carefully defined classes and
relations.  This suggests that subclasses and subrelations with rules
inter-relating them are needed.    (044)

  (genls MonogomousMarriage Marriage)
  (genls TraditionalMarriage Marriage)
Depending on the definition of TraditionalMarriage:
  (genls TraditionalMarriage MonogomousMarriage)    (045)

  (genls ArrangedMarriage Marriage)
  (genls PolygomousMarriage Marriage)
  (genls PolyamorousMarriage PolygomousMarriage)
  (genls SameSexMarriage Marriage)
  (genls InterracialMarriage Marriage)
  (genls ChildMarriage Marriage)    (046)

  (disjointWith MonogomousMarriage PolygomousMarriage)
  (disjointWith TraditionalMarriage PolyamorousMarriage)
  (disjointWith TraditionalMarriage SameSexMarriage)
Depending on the definition of TraditionalMarriage:
  (disjointWith TraditionalMarriage InterracialMarriage)    (047)

  (isa spouseInMarriage BinaryPredicate)
  (arg1Isa spouseInMarriage Person)
  (arg2Isa spouseInMarriage Marriage)
  (relationAllExistsMin spouseInMarrage Marriage 2)
  (relationAllExistsCount spouseInMarrage MonogomousMarriage 2)    (048)

  (relationAllExistsType spouseInMarrage MonogomousMarriage HumanMale 1)
  (relationAllExistsType spouseInMarrage MonogomousMarriage HumanFemale 1)    (049)

  (relationAllExistsCount initialSubEvent Marriage Wedding 1)
  (isa spouseOf CoEquivalenceBinaryPredicate)    (050)

  (isa husbandOf IrreflexiveBinaryPredicate)
  (genlPreds husbandOf spouseOf)
  (arg1Isa husbandOf HumanMale)
  (arg2Isa husbandOf Person)
  (equivPredsWRTTypes husbandOf spouseOf HumanMale HomoSapiens)    (051)

  (isa wifeOf IrreflexiveBinaryPredicate)
  (arg1Isa wifeOf HumanFemale)
  (arg2Isa wifeOf Person)
  (genlPreds wifeOf spouseOf)
  (equivPredsWRTTypes wifeOf spouseOf HumanFemale HomoSapiens)
  (negationPreds wifeOf husbandOf)    (052)

        (spouseInMarriage ?MARRIAGE ?SPOUSE1)
        (spouseInMarriage ?MARRIAGE ?SPOUSE2)
        (different ?SPOUSE1 ?SPOUSE2))
     (holdsDuring ?MARRIAGE
         (spouseOf ?SPOUSE1 ?SPOUSE2)))
        (isa ?MARRIAGE TraditionalMarriage)
        (spouseInMarriage ?MARRIAGE ?HUSBAND)
        (spouseInMarriage ?MARRIAGE ?WIFE)
        (isa ?HUSBAND HumanMale)))
        (different ?HUSBAND ?WIFE))
     (holdsDuring ?MARRIAGE
         (isa ?WIFE HumanFemale)))
        (isa ?MARRIAGE ChldMarriage)
        (subEvents ?MARRIAGE ?WEDDING)
        (isa ?WEDDING Wedding)))
        (different ?HUSBAND ?WIFE))
     (holdsDuring ?WEDDING
        (thereExists ?SPOUSE
            (spouseInMarriage ?MARRIAGE ?SPOUSE)
            (isa ?SPOUSE HumanChild)))))
        (spouseOf ?HUSBAND ?SPOUSE2)
        (isa ?HUSBAND HumanMale))
     (husbandOf ?HUSBAND ?SPOUSE2))
        (spouseOf ?WIFE ?SPOUSE2)
        (isa ?WIFE HumanFemale))
     (wifeOf ?WIFE ?SPOUSE2))
        (isa ?MARRIAGE SameSexMarriage)
        (spouseInMarriage ?MARRIAGE ?SPOUSE1)
        (spouseInMarriage ?MARRIAGE ?SPOUSE2)
        (isa ?SPOUSE1 HumanMale)
        (different ?SPOUSE1 ?SPOUSE2))
        (isa ?SPOUSE2 HumanMale))
        (isa ?MARRIAGE SameSexMarriage)
        (spouseInMarriage ?MARRIAGE ?SPOUSE1)
        (spouseInMarriage ?MARRIAGE ?SPOUSE2)
        (isa ?SPOUSE1 HumanFemale)
        (different ?SPOUSE1 ?SPOUSE2))
        (isa ?SPOUSE2 HumanFemale))    (053)

>> We should also be able to relate very common concepts
>> like "currently" to anything that may have a time dimension and not
>> imbed such concepts in these made- up phrases.    (054)

Agreed.  The issue of lack of temporal assertion on predicates becomes
an issue, for example when assertions about serial monogamous marriages
conflict for rules about monogamous marriages.    (055)

>> (I have no problem with making
>> real composite concepts that are compositions, but we should then
>> understand their parts).  To do this we much be able to use the
>> "relations" (of any granularity) as subjects of such relations.    (056)

>> If there is a proper representation of the terms and concepts of a
>> complex
>> concept like marriage these can be properly related.  "Married",
>> "Spouse",
>> "Husband", "Wife", etc. are all only meaningful within the context of
>> marriage.  Expressing the verb phrase "x isMarriedTo y" or "x hasSpouse
>> y"
>> or "Marriage(x,y) are saying the SAME THING, it is the same fact
>> expressed
>> in a different way - in some cases as relations and some cases reified.    (057)

None of these relations references the marriage itself.  If there have
been multiple marriages between the same two people, it could not be
expressed using any of these predicates.    (058)

>> I
>> submit that to understand how the different ways we express information
>> are related we have to understand concepts at this level.    (059)

IMHO, it is difficult to understand such different ways of understanding
information without reifying situations such as the marriage.    (060)

-- doug f    (061)

>> -Cory
>> -----Original Message-----
>> From: doug foxvog
>> Sent: Saturday, November 05, 2011 11:24 AM
>> On Fri, November 4, 2011 14:02, Cory Casanave said:
>>> The other strong use-case for reification, besides n-ary, is to
>>> support relations as first-class elements that can also be the subject
>>> of other relations.  I have found this essential to represent the
>>> concepts of a domain accurately - "marriage" is such a relation.    (062)

>> In ontological terms, Marriage is a temporal situation.
>> "isCurrentlyMarriedTo" is a relation -- in this case a binary relation.
>> Beginning ontologists often start creating binary and multiple arity
>> relations to represent sets of columns in a database, not stopping to
>> consider what the underlying classes of things are and realizing that
>> many more relations could apply to those classes of things in various
>> circumstances.  Events and situations are common categories of things
>> that are often so modeled.
>> ...
>> -- doug foxvog
>> ...    (063)

>>> -Cory
>>> -----Original Message-----
>>> From: Ed Barkmeyer
>>> Sent: Friday, November 04, 2011 12:32 PM    (064)

>>> The practice of reifying relations in binary models goes back at least
>>> to Peter Chen and the original Entity-Relationship models.
>>> That is, you make the relation itself a 'class'/'entity', and then it
>>> has binary relationships to each of its arguments.  ...    (065)

>>> This is precisely the recommended best practice for representing n-ary
>>> relationships in OWL:  the relation becomes a 'class', and each of the
>>> argument slots becomes an objectProperty (or datatypeProperty) named
>>> for the role.  ...    (066)

>>> -Ed
>>> --
>>> Edward J. Barkmeyer                        Email: edbark@xxxxxxxx
>>> National Institute of Standards & Technology Manufacturing Systems
>>> Integration Division
>>> 100 Bureau Drive, Stop 8263                Tel: +1 301-975-3528
>>> Gaithersburg, MD 20899-8263                Cel: +1 240-672-5800
>>> "The opinions expressed above do not reflect consensus of NIST,  and
>>> have not been reviewed by any Government authority."
>>> ...    (067)

doug foxvog    doug@xxxxxxxxxx   http://ProgressiveAustin.org    (068)

"I speak as an American to the leaders of my own nation. The great
initiative in this war is ours. The initiative to stop it must be ours."
    - Dr. Martin Luther King Jr.
=============================================================    (069)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (070)

<Prev in Thread] Current Thread [Next in Thread>