John,
I wasn’t recommending particular identifiers for the relations. I just used those as examples that would suggest the intended meaning. I noted at the end
that ISO 15926 defines more abstract relations that can be used for the same purposes. You are certainly entitled to use whatever terms you please.
-Ed
From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx]
On Behalf Of John McClure
Sent: Wednesday, February 26, 2014 1:01 PM
To: [ontolog-forum]
Subject: [ontolog-forum] SituationInvolvesRole (was Defining everything in terms of relations)
Instead of "SituationInvolvesRole"
how about terms like "may-be:in" or "may-have:some" or "may-have:a" to relate concepts?
Here's an example of the verb:article /
verb:preposition methodology I've spoken of, in a model for 'newspaper' that I put together the other day that I think demonstrates how accessible this methodology can be.
thanks/jmc
On 2/26/2014 8:37 AM, Barkmeyer, Edward J wrote:
Pat,
You wrote:
> You can treat each ‘Role’ as a relation with a ‘range’, but one must never confuse a (non-unary) relation with a class (type). But One can alternatively
have roles as types (not as relations), in which case one would have a different relation to relate some Object instance with its ‘Role” type.
If I understand this correctly, I agree.
If, as William suggests, you model the ‘domain verbs’ as classes of states/activities, then a ‘role instance’ is a pair: (state/activity instance, role player),
and the Role is a binary relation that those pairs satisfy. I think this is what you mean by “treating each Role as a relation”.
The alternative is that a Role is a ‘type’ whose instances are ‘things in the process of playing the role’, i.e. “participations”. Each ‘role instance’ is
an instance of the Role type. Some binary relation, e.g., SituationInvolvesRole, relates each ‘role instance’ to the state/event in which the Role is instantiated/played. That is, the instances of SituationInvolvesRole are pairs of the form (state/activity
instance, role instance). It is a necessary property of role instances that they participate in exactly one such pair. Further, each role instance is related to the object that plays the Role in that state/activity by a different relation ThingPlaysRole,
and its instances are pairs of the form (role player, role instance). I understand this to be what you mean by your second option above.
Note that the mulitiplicity rule for participation of a role instance in SituationInvolvesRole and ThingPlaysRole pairs may depend on the role itself, and the
modeler’s view. One can consider that there might be states/activities that have ‘optional roles’ – roles that may or may not be played in various instances of the ‘situation type’ that defines the roles. Further, there may be multiple undistinguished roles
in a state/activity, so the same Role (type) could have multiple role instances in the same state/activity. One might, I suppose, alternatively model the latter as one role instance that is related to multiple role players via ThingPlaysRole. (I would strongly
prefer multiple role instances of the same Role type to multiple players of the same role instance, but I think that is some “intuitive ontological commitment”. I don’t know that multiple players of the same role instance would produce problems; I just “would
not do that”.)
Note also that, in either case above, the Role can have refinements, e.g. captain of ship is a narrower role than officer on ship. If you model the officer
role as a relation (Pat’s first option), e.g., an OWL objectProperty, then the captain relation is a owl;subpropertyOf the officer property. Similarly, if you model the officer Role as a type/class (Pat’s second option), then the captain role is a owl:subclassOf
the officer class.
Final Note: Pat’s proposed Role class, e.g., “being an officer”, has instances that are the participations in states/activities; its instances are not the
role players themselves. The class Officer defined by “person who is an officer on some/any ship” has instances that are the role players themselves. They are significantly different classes.
Pat wrote:
> By now I would hope that discussants on this list would be careful not to imply that there is some necessary formal meaning for any linguistic term. In an
ontology, the entity labels mean only what the logic says they mean.
I fully agree. William’s point was that, in addition, there is no necessary formal model for linguistic parts of speech: A verb can be a Class, a noun can
be a Property. Put another way, when you title an OWL Class ‘Person’, ‘Person’ is NOT an English word; it is an identifier/term in the OWL language, which is radically different from English.
(In the early 1990s, I railed in vain against an ISO TC184/SC4 modeling dictum that did not permit the use of the English word ‘product’ in the definition of
the EXPRESS symbol ‘product’ because “the definition would be circular”. That rule demonstrated that the people making the rules did not understand what they were doing.)
-Ed
P.S. According to Matthew West (separate discussion), the ISO 15926 ontology makes the ‘role instances’ a class of ‘temporal parts’ of the role players, using
TemporalPartOf for ThingPlaysRole; similarly, it makes the role instances ‘spatial parts of’ the situation objects, using SpatialPartOf for SituationInvolvesRole (with the inverted ‘reading’). So the 15926 approach matches Pat’s second option.
One point regarding Ed Barkmeyer’s comment:
[EB] > Right! This is even worse than the problem of making the role a subtype of its ‘range’:
> the things that can play the role.
You can treat each ‘Role’ as a relation with a ‘range’, but one must never confuse a (non-unary) relation with a class (type). But One can alternatively
have roles as types (not as relations), in which case one would have a different relation to relate some Object instance with its ‘Role” type.
I actually use the latter formalism in COSMO, and I don’t find any logical inconsistencies. Perhaps you have in mind some examples where this creates some
problem.
[EB} > This makes the range type a subtype of ‘role’,
> which is properly not a ‘type’ at all.
Well, in the COSMO treatment that is not implied. In COSMO the type ‘Role’ is a top-level type, and many things such as ‘Student’ can be represented as
both a subtype of ‘Person’ (PhysicalObject) and of ‘Role’ (a time-dependent entity). This means that an instance of ‘Student’ can only be used in an assertion when the valid time interval of the assertion is within the time interval of that ‘Student’ role.
The general class of ‘HumanRole’ is used as the parent class for all Roles filled by humans, and ‘Student’ and ‘Teacher’ are a subtypes of ‘HumanRole’. Every subtype of ‘HumanRole’ is a subtype of both Human and Role.
As for the comment that a ‘Teacher’ can be a computer, in COSMO an object of that type would not be an instance of class ‘Teacher’, but would be represented
as an instance of some subtype of computer program or computer or some combination thereof (depending on what you intend to refer to). It can be related to the human role by some relation such as ‘performsTheFunctionOfa’. Of course, one can, in informal
speech, use any label for any entity, but in the Ontology the term labels are unique with unique meanings. The ontology type ‘Student’ or ‘Teacher’ in COSMO can only refer to humans within the time period when they are serving ion those roles. In informal
speech, they can refer to anything (e.g. ‘experience is the best teacher’).
By now I would hope that discussants on this list would be careful not to imply that there is some necessary formal meaning for any linguistic term. In an
ontology, the entity labels mean only what the logic says they mean.
This COSMO ‘Role’ formalism is adopted for a couple of reasons: to keep the ontology as close as possible to linguistic usage (we say that ‘the student
went to class’, not ‘the person who was playing the role of student went to class’), and to allow the most convenient representation within the OWL formalism. For many purposes, the ontology will need to be converted into FOL, but I believe that there are
good reasons to have an OWL version as well as an FOL version, so the first version of COSMO is being developed in OWL, with the goal of being as perspicuous as is consistent with logical consistency. Converting from OWL to FOL will be the easier path to
having both formalisms available.
The general problem I have with discussions such as these is that they are occurring outside of any specific example application that can provide a test
or at least some constraints on proposals for the utility of some formalism. I would like to see more examples from practical uses to help decide what formalism is useful in what situation.
Pat
I intended to add this to my previous email.
William wrote:
> Nothing wrong with the word 'attribute', or the way CHEN and now
> everyone uses it.
>
> EXCEPT FOR Tthe BIG MISTAKE of thinking that a certain kind of thingie
> (say blue) IS an attribute, just because it **can play that role.**
Right! This is even worse than the problem of making the role a subtype of its ‘range’: the things that can play the role. This makes the range type a subtype
of ‘role’, which is properly not a ‘type’ at all.
I think the origin of this error is the specific distinction in Chen’s model between ‘attributes’, as relationships to ‘values’ (computational representations),
and ‘relationships’ as relationships to ‘entities’. If you view the ‘value’ as JUST the computational entity – an integer, a real number, a character string -- then the ‘role’ actually gives it useful semantics. It is not that ‘blue’ IS an attribute of things,
but rather that the string “blue” AS an attribute refers to a color of things.
In Sjir Nijssen’s approach, by comparison, Color is a ‘non-lexical object type’ and Blue is an instance of that type, but, when converting that model to a database
schema, we may decide that the “NOLOT” Color can be represented by a ‘lexical object type (LOT)’ that is said to be a subtype of String. This sorts out the distinction between the semantic Blue and the string “blue”. In a similar way, the ISO 11179 metadata
model distinguishes between the ‘conceptual value’ Blue and the ‘data element value’ “blue”.
But we can think of this as a progression of understanding – Chen (1976), Nijssen (1982), ISO 11179 (1998).
-Ed
On Mon, Feb 24, 2014 at 11:10 PM, John F Sowa <sowa@xxxxxxxxxxx> wrote:
Ed, Pat C, William,
.....
WF
> Nothing wrong with the word 'attribute', or the way CHEN and now
> everyone uses it.
>
> EXCEPT FOR Tthe BIG MISTAKE of thinking that a certain kind of thingie
> (say blue) IS an attribute, just because it **can play that role.**
I agree. I would use a dyadic relation named HasAttr to relate
an entity to something called an attribute of that entity.
An attribute of something is only an attribute when it is viewed
in the role of the second argument of a relation named 'HasAttr'.
Right, I have been saying the following on this forum for almost two years:
being an attribute is, in human language, not a fixed feature of a word or a concept, except for some of the words in some minority of languages. In others, 'beauty' and 'beautiful' and 'is beautiful' are expressed with exactly the same word, and the role
in the sentence is market either by word position (in a positional language) or a particle that shows this (in a tagged language).
In a very mongrel language like English, things are very complex, because multiple patterns are in play at the same time, on a word-by-word basis. For example, some words, like 'German', and the same as nouns and adjectives. He is German, He is A German, while
others that occupy the same semantic space, like English and Scottish, are not the same, but have to turn into English Person and Scot.
psycholinquists are examining how various language-related disabilities are manifested by speakers of different languages, and thus finding more about what is the same and different about the fundamental samenesses
and differences between thinking and speaking in different languages. They don't use entities, attributes, and relationships as their foundation, nor parts of speech like nouns and verbs. They tend to use theta roles.
Given this, why would one want to insist that the fundamental organization of thought or a universal simple way to express propositions is based on entities, their attributes, and their relationships?
> In another attibutive relation, the same thingie, blue, can play
> the role of attributed to, such as in 'blue is a color.'
This raises the question about instances of blueness. We'd like to say
the blue of your coat is darker than the blue of the sky. So we need
to include instances of blueness in the ontology: the HasAttr relation
relates a coat x to an instance of blueness y, which we can relate to
another instance of blueness z of the sky.
Yes, this is yet another role that 'blue' can plan. It can be the name for a particular repeatable experience of blueness.
The blue of my coat is, if I understand correctly, what some metaphysicians are calling 'tropes' these days.
" According to trope theory, the world consists (wholly or partly) of ontologically unstructured (simple) abstract particulars or, as they are normally called,
tropes. " Stanford Encyclopedia of Philosophy
That is, it is a particular but repeatable experience of blue. And, in fact, if I understand, in this thought, 'my coat' is there just to locate the blue experience using a shared
identifier, becasue the experience of that blueness, along with many other cues, is in fact one of the things that underlies constructing the coat in question out of our fundamental experiences. Personally, I like this direction for metaphysics, in that
it brings it closer to psychophysics and social psychology.
WF
> Did you ever have clients who had trouble seeing that 'customer'
> was a role, and that so was 'vendor,' so that the vendor and customer
> might be the same company?
Good old Aristotle was quite clear about those issues. He
distinguished the "substance" -- such as HumanBeing -- from the
"accidents" such as being blue or being a customer, vendor, etc.
For an Aristotelian description of George Washington, see slide 6
of http://www.jfsowa.com/talks/aristo.pdf
Yes, and this goes back to some previous discussions here about natural types, with which I also agree.
But I have found it harder to sell that idea right off the bat than selling the understanding that some words used to describe things are actually describing roles of the things (customer, scoutmaster) , and others are not (person).
John
_________________________________________________________________
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
|
_________________________________________________________________
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 (01)
|