RE: SUO: RE: Re: Semiotics Formalization
Dear Adam,
See responses below marked MW:
Regards
Matthew
============================================
Matthew West
Asset Information Management
Shell Services International
H3229, Shell Centre, London, SE1 7NA, UK.
Tel: +44 207 934 4490 Fax: 7929
Mobile: +44 7796 336538
E-mail: Matthew.R.West@is.shell.com
http://www.shellservices.com/ <http://www.shellservices.com/>
============================================
-----Original Message-----
From: apease [mailto:apease@teknowledge.com]
Sent: 20 September 2000 22:45
To: West, Matthew MR SSI-GREA-UK; Jon Awbrey; Stand Up Ontology
Subject: RE: SUO: RE: Re: Semiotics Formalization
Matthew,
Comments below
At 07:19 PM 9/19/2000 +0200, West, Matthew MR SSI-GREA-UK wrote:
Dear Adam,
See comments marked MW: Below.
Regards
Matthew
============================================
Matthew West
Asset Information Management
Shell Services International
H3229, Shell Centre, London, SE1 7NA, UK.
Tel: +44 207 934 4490 Fax: 7929
Mobile: +44 7796 336538
E-mail: Matthew.R.West@is.shell.com
http://www.shellservices.com/ <http://www.shellservices.com/> <
http://www.shellservices.com/ <http://www.shellservices.com/> >
============================================
-----Original Message-----
From: apease [ mailto:apease@teknowledge.com <mailto:apease@teknowledge.com>
]
Sent: 19 September 2000 00:36
To: West, Matthew MR SSI-GREA-UK; Jon Awbrey; Stand Up Ontology
Subject: Re: SUO: RE: Re: Semiotics Formalization
Matthew,
I'd like to relate your work to the terms I sketched earlier. There's
more than one way to formalize the same concepts so I'll just state my
preferences. Hopefully, discussing preference for how formalizations are
done may help the SUO in addition to the conclusions we may arrive at in
terms of particular ontology content.
Comments below:
At 08:34 PM 9/15/2000 +0200, West, Matthew MR SSI-GREA-UK wrote:
Dear colleagues,
We (EPISTLE) recognise elements of information at 3 basic levels.
Level 1 - information carrier
Individuals that carry some pattern that can be decoded are known as
information carriers (this is not what we call a generic type, i.e.
reflecting the underlying nature of the individual).
Examples include:
- a paper document with writing on it.
- a hard disc in a computer with magnetic marks on it
- the dance of a bee to show its hive where honey can be found
- the DNA in a molecule.
An example of a pattern that does not contain information:
- the pattern made by the cement and bricks in a wall.
Note that the pattern does not have to be human interpretable, but it does
have to be capable of interpretation.
I think that virtually any physical thing can carry information. The
ability to carry information is not part of the identity of the object.
MW: This is what I meant by saying that information carrier was not a
generic type.
ok
> So, I'd suggest that this notion should be coded as a relation between the
physical thing and the information it carries. In the list I proposed
earlier this would be 'realization-of'.
> As an example, we might have
> (instance-of MyCopyOfHamlet Book)
> (instance-of Hamlet Play-Theater)
> (realization-of MyCopyOfHamlet Hamlet)
MW: I would also see the realization-of relation above as a classification.
I'm not sure what you mean.
MW: Let me try to explain. Hamlet here is a class that has a specification,
and what you want is something that meets that specification, i.e. is a
member of the class. Your copy of Hamlet is an object that meets the
specification. A specification is that basis for inclusion and exclusion for
the membership of the class. Realisation also has some aspect of creation
about it, and I would say could be decomposed into the simpler concepts of
creation of object and classification of object.
Level 2 - class of encoded information
The pattern conveyed by the information carrier. This is an abstract object
(we would call it a class of individual - where members of the class have a
common pattern). So for example, if you had 5 copies of a paper document,
the class of encoded information is the pattern that they share.
Examples include:
- the words on a page of text
- the ASCII codes that can be derived from reading a hard disk
Note: there can be transformations between encodings.
If I understand your explanation here, this would be equivalent to what I'm
calling 'Information'. This notion subsumes your Level 3 encoding. This
could include subclasses for the concepts expressed by words as well as the
expression of those concepts as English and the expression of that English
as symbols on a printed page. I've offered 'coding-of' as a relation which
connects one instance of information with another instance that codes the
same information in a different way.
MW: I'm not sure I understand what you (or others) mean by "subsumes" (even
after looking it up in the OED). I would appreciate an explanation.
By subsumes I meant, broadly, covers or includes. However, on second
reading it appears to me that you may have intended a more specific meaning
for your level three akin to a class of logical forms, whereas I intended
information to cover any sort of coding.
MW: For me information means "that which is encoded independent of any
particular encoding"
M W: I think "translation" or "transformation" is better than "coding of" as
the relation between different classes of encoded information.
Hmm. Good names are hard to find. 'translation' has the connotation to me
of converting between human languages (true, there are other definitions
but...). 'transformation' has the implication (for me at least) of covering
physical changes.
MW: This may be how you are used to using translation, but it is also used
for e.g. computer based languages. Another word that might be suitable is
conversion.
Level 3 - class of information
meaning independent of how it can be encoded.
e.g. the idea/concept of "Matthew", independent of whether it is e.g. spoken
or written.
At each of these levels information can be classified: e.g.
Classes of Information Carrier
- paper document
- hard disk
Class of class of encoded information
- ASCII
- English
Class of class of information
- entity type
- relation
- predicate
I'd propose that we have an information carrier relation, a class of
information for concepts and a class for various expressions of concepts.
MW: This seems close to what I would propose, but I believe the information
carrier relation you propose is really simply a classification relation, and
that even though information carrier is a non-essential class of individual,
it is a useful one.
By this do you mean that you still want to reify the class
'InformationCarrier' in the ontology?
MW: Is it worth having as a recognised concept? yes. Is it an essential
class? no. Is it derived from the relations an object has? yes (but then so
int he end aare almost all classes, including things like Person).
Regards
Matthew
============================================
Matthew West
Asset Information Management
Shell Services International
H3229, Shell Centre, London, SE1 7NA, UK.
Tel: +44 207 934 4490 Fax: 7929
E-mail: Matthew.R.West@is.shell.com
http://www.shellservices.com/ <http://www.shellservices.com/> <
http://www.shellservices.com/ <http://www.shellservices.com/> >
============================================
> -----Original Message-----
> From: Jon Awbrey [ mailto:jawbrey@oakland.edu <mailto:jawbrey@oakland.edu>
< mailto:jawbrey@oakland.edu <mailto:jawbrey@oakland.edu> >
]
> Sent: 14 September 2000 22:44
> To: Stand Up Ontology
> Subject: SUO: Re: Semiotics Formalization
>
>
>
> Adam Pease (in reply to an off-list messaage) wrote:
> >
> > Jon,
> >
> > I appreciate your comments but I'm not quite sure how they
> > add to the ontology. Are you proposing a Sign class which
> > is a subclass of Information and subclass of PhysicialThing?
> >
> > Adam
> >
> <...>
>
> ¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤
>
> Adam,
>
> I do not know how others understand the word "information",
> but for me it has its meaning in a particular kind of context.
>
> In the setting where this notion of information makes sense to me,
> we have, as a part of the overall background set-up, a measure of
> a quantity called "entropy" or "uncertainty". This is a measure on
> distributions ("frequency distributions" or "probability densities").
>
> The distribution F : X -> R is a function from a sample space X
> to the real numbers R, such that the real value F(x) EUR R can be
> interpreted as the probability that x will happen. (More carefully,
> this interpretation only works for discrete frequency distributions,
> but that is more or less the general idea.) (Also, it is traditional
> to use a capital omega for the sample space, for "outcomes" or perhaps
> for "occurrences", I think, but here I will have to sign it
> with an "X".)
>
> The entropy or uncertainty measure M is thus a function of the
> type M : (X -> R) -> R, and it satisfies some additional axioms
> that make it a decent formalization of our intuitive notions of
> doubt or uncertainty in the situations that are described by the
> distributions of type (X -> R).
>
> In this setting, if we can say what our measure of uncertainty
> would be both before and after receiving a particular sign, say,
> by way of making an observation or by way of some other courier,
> then we can define a quantity called the "information capacity"
> of the set of signs at issue. The information capacity is the
> "average uncertainty reduction on receiving each sign" (AURORES).
> Thus comes the dawn of information theory.
>
> The way I understand it, sign-tokens are physical things --
> they are actually another set of "outcomes" that "occur"
> in the real world, and so they have to have some sort of
> physical basis, but sign-types are classes of sign-tokens,
> what statistical folks called "events", that is, subsets
> of some sample space, and so they have an abstract quality
> to them. We "typically" intend the token as a representative
> of its class, and so there is almost always some ambiguity here.
>
> So when I hear people talking about a category of being that
> they call "information", I have to stop and say to myself:
> Okay, they mean a sign that is given in a setting where it
> posseses and potentially conveys a quantity of information.
>
> Anyway, this is how I understand these things. As I read over what
> I just wrote, however, I notice that it looks more like a personal
> attempt to come up with a rational -- and risky -- reconstruction
> of what I rememeber of classical information theory, but within the
> kind of sign-theoretic setting that I get from Peirce, and I now see
> that many of the details are looking kind of fuzzy.
>
> So, thanks for the chance to at least start on the work of
> clarification,
> and I will shift this back to on-list communication so as not
> to protect
> it from the criticism of others who might be able to help out here.
>
> Regards,
>
> Jon
>
> ¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤~~~~~~~~~¤
>
-----------------
Adam Pease
Teknowledge
(650) 424-0500 x571
-----------------
Adam Pease
Teknowledge
(650) 424-0500 x571