[Top] [All Lists]

Re: [ontolog-forum] ambiguity interferes with

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: Patrick Durusau <patrick@xxxxxxxxxxx>
Date: Mon, 12 Mar 2007 08:38:21 -0400
Message-id: <45F549BD.5030904@xxxxxxxxxxx>
Deborah,    (01)

Deborah MacPherson wrote:    (02)

> Right - but - when you are drawing, for example in AutoCAD, there are 
> however many underlying strings of codes, positions, and lengths 
> relative to an arbitrary 0,0....the act of drawing simply happens to 
> figure out relationships in a methodical fashion and the operators 
> never see the bottom level the computer uses to capture or regenerate 
> the drawing. There is no reason for a CAD operator to look at at that 
> level to see what they are drawing. Adrian Walker shows this as a blue 
> line - tasks and decisions that happen above or below the line.
I assume you mean the "blue line" on Adrian's slide 28 for example?    (03)

I am not sure that is a helpful distinction other than to illustrate 
that most users don't consider that there may be a semantic disconnect 
between their use of a term and the term used by others. Adrian says 
"model" and I suppose that is fair since neither users uses a term in 
isolation from all the other terms in their environment.    (04)

> Where this distinction occurs in the subject maps you refer to are 
> what I will be looking for in the link below. Honestly, I don't care 
> about naming, I prefer numbers. I only really care about proportions 
> and putting things that belong together next to each other for easy 
> access and a flow or planned unfolding of a story - as if conceptual 
> or digital structures could be like visiting a museum and feel like 
> you are discovering and seeing but actually being led through by 
> someone who loves the materials.
Well, it is certainly possible to construct a subject map that results 
in the type of goo that is the result of a Google search. Nothing in 
that paradigm guarantees that authors are going to make sensible or even 
useful choices. I suspect that is rather difficult for any abstract 
formal model to enforce.    (05)

What subject maps do recognize is that you and I may have completely 
different ways to identify the same subject. What is more, either one of 
us or even a third person, take Jack in this case, can create a rule 
that adds the same third identification to both of our proxies along 
with a rule that says in the presence of that third identification, 
these proxies merge. (That is one model of merging, there are others.) 
The result is a single representative for the same subject.    (06)

Note that each of our identifications persist (at least under the model 
I have suggested), which means that we have not arrived at a common 
identification (one of the failing goals of the SW) and that we can in 
fact find the "foreign" identifications when we arrive at the proxy by 
our identifications.    (07)

While it is possible in some situations to write inference rules to 
result in that sort of merging, what puzzles me is the apparent 
adversion to using the currently best inference engine that is 
available, people. Granted there is a lot of data already and more 
appearing everyday but surely the authors of that data thought it meant 
something in some particular domain.    (08)

In a following post you make the point that simply knowing the domain 
would be a big step in terms of dealing with ambiguity. Take gene names 
for example. Simply knowing that a term is used for a gene would at 
least reduce the problem to one of ambiguity in the domain of gene 
names. And there are distinctions within that domain that are almost as 
simple and that would further reduce the ambiguity. All without 
constructing ontologies or other mechanisms to assist in the research 
problem of "intelligent" agents. Not that I am unsympathetic to that 
problem, but I would rather have useful results sooner than later. Or in 
the case of "intelligent" agents as envisioned by TBL, much later (if ever).    (09)

Hope you are at the start of a great week!    (010)

Patrick    (011)

> Debbie
> On 3/11/07, *Jack Park* <jack.park@xxxxxxx <mailto:jack.park@xxxxxxx>> 
> wrote:
>     Debbie,
>     There is a bit more at
>     It's worth reiterating one point in light of your inquiry: subject
>     proxy
>     objects  are the objects contained in a subject map. A subject proxy,
>     itself, could be as simple as a database table the columns of
>     which are
>     propertyType and propertyValue (in the simplest possible rendition).
>     Therefore, RDF is a reasonable way to represent the properties
>     contained
>     in a subject proxy.
>     Among those properties will be "name properties"; every knowable name
>     for the subject represented by the subject proxy can be gathered
>     together for each with some metadata such as language or usage or name
>     type (e.g. acronym). With that organization, you can create a social
>     bookmarking website such as Tagomizer or Fuzzzy and, at the same
>     time,
>     have a name be used without ambiguity in relation to many different
>     subjects. Some names, e.g. "George Bush" can serve the proxies that
>     represent both presidents and a tag with the same name. Each actual
>     proxy carries a locator, the database identity, which, itself,
>     could be
>     a URI as has been discussed at length here. By separating names from
>     identity, we are able to accommodate many uses for the same name
>     string.
>     Other properties facilitate disambiguation when a name string is
>     used as
>     a query.
>     IMHO, with that architecture, or at least attention to what that
>     architecture offers as one creates ontologies, conceptual graphs, or
>     other representation systems, one need not write any "design criteria"
>     or even "building codes" for naming since the identification scheme is
>     decoupled from naming conventions. Within subject/object
>     identification
>     schemes, I have no doubt that "building codes" will be appropriate.
>     Please note that I am using the term "naming conventions" here in
>     a way
>     that seems vastly different from the way everybody else appears to
>     use
>     that term; indeed, I suspect that when the term "name" is applied to a
>     concept, what is meant is "identifier". I use the term "naming
>     convention" more as an oxymoron: the only naming conventions I have
>     experienced appear in such contexts as, for example, giving the first
>     daughter the grandmother's name as a middle name, following the
>     appending of "jr", "II", etc to children who take a parent's name,
>     and
>     other naming practices. I'm waving arms and leaving a lot out
>     here, but
>     my intent is to show how I see "naming conventions" in the grand
>     scheme
>     of communications.
>     In terms of naming subjects, well, those come from everywhere and are
>     just strings that are often ambiguous, as my experience with homeland
>     security suggests.
>     In terms of identifying conventions, the story is quite different;
>     names
>     *can* serve in identification, but they often should not be the sole
>     property associated with identity. My experience with Philippine
>     Airlines giving my frequent flyer mileage to someone who was born on a
>     different date strongly suggests their computer programmers don't
>     know that.
>     Jack
>     Deborah MacPherson wrote:
>     > Thanks John, printing and reading. Actually, I think ambiguity
>     may be
>     > what makes the world go around and look forward to reading these
>     materials.
>     >
>     > Thanks Jack, I'm so happy to see you describe the examples as
>     "crafted"
>     > and will look at these. It would be interesting to hear the
>     > author/artist Steve Newcombs views. I realize now after reading
>     > ontolog-forum that probably homegrown tags and tag gardeners are
>     what
>     > I'm interested in mainly to connect these people to their
>     counterparts
>     > at museums and other government organizations with public
>     information.
>     >
>     > In my opinoin, what needs to happen is writing or specifying
>     this design
>     > criteria, which I don't think is an ontology after all. If there
>     could
>     > be what Pat Hayes called a kind of "building code" that would
>     really let
>     > the tag gardeners and PhD curators, research scientists and others
>     > actually connect to each other on a level playing field. Will
>     look at
>     > your suggested links and digest them for awhile now. Thanks.
>     >
>     > This things have been bothering me and I appreciate the pointers
>     and
>     > explanations.
>     >
>     > Sincerely,
>     >
>     > Debbie
>     >
>     > On 3/10/07, *John F. Sowa* < sowa@xxxxxxxxxxx
>     <mailto:sowa@xxxxxxxxxxx> <mailto: sowa@xxxxxxxxxxx
>     <mailto:sowa@xxxxxxxxxxx>>>
>     > wrote:
>     >
>     >     Debbie,
>     >
>     >     Ambiguity is inevitable.
>     >
>     >     It sometimes causes trouble.  For certain special cases,
>     >     it is possible to create artificial languages that
>     >     eliminate ambiguity -- but *only* for those special cases.
>     >
>     >     It is not possible to eliminate ambiguity, because that
>     >     would also freeze science, engineering, business, art, and
>     >     *life* in a static, unalterable toy universe that endlessly
>     >     repeats its preset moves -- like a computer program.
>     >
>     >     See the following abstract and paper.
>     >
>     >     John
>     >     ________________________________________________
>     >
>     >     Source: http://www.jfsowa.com/pubs/dynonto.htm
>     >
>     >     A Dynamic Theory of Ontology
>     >
>     >     John F. Sowa
>     >
>     >     Abstract.  Natural languages are easy to learn by infants,
>     they can
>     >     express any thought that any adult might ever conceive, and they
>     >     accommodate the limitations of human breathing rates and
>     short-term
>     >     memory.  The first property implies a finite vocabulary, the
>     second
>     >     implies infinite extensibility, and the third implies a
>     small upper
>     >     bound on the length of phrases.  Together, they imply that
>     most words in
>     >     a natural language will have an open-ended number of senses
>      ambiguity
>     >     is inevitable.  Peirce and Wittgenstein are two philosophers who
>     >     understood that vagueness and ambiguity are not defects in
>     language, but
>     >     essential properties that enable it to accommodate anything and
>     >     everything that people need to say.  In analyzing the
>     ambiguities,
>     >     Wittgenstein developed his theory of language games, which
>     allow words
>     >     to have different senses in different contexts,
>     applications, or modes
>     >     of use.  Recent developments in lexical semantics, which are
>     remarkably
>     >     compatible with the views of Peirce and Wittgenstein, are
>     based on the
>     >     recognition that words have an open-ended number of
>     dynamically changing
>     >     and context-dependent microsenses. The resulting flexibility
>     enables
>     >     natural languages to adapt to any possible subject from any
>     perspective
>     >     for any humanly conceivable purpose. To achieve a comparable
>     level of
>     >     flexibility with formal ontologies, this paper proposes an
>     organization
>     >     with a dynamically evolving collection of formal theories,
>     systematic
>     >     mappings to formal concept types and informal lexicons of
>     natural
>     >     language terms, and a modularity that allows independent
>     distributed
>     >     development and extension of all resources, formal and informal.
>     >
>     >
>     _________________________________________________________________
>     Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
>     Subscribe/Config:
>     http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
>     <http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/>
>     Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
>     <mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx>
>     Shared Files: http://ontolog.cim3.net/file/
>     Community Wiki: http://ontolog.cim3.net/wiki/
>     To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>     <mailto:ontolog-forum@xxxxxxxxxxxxxxxx>
> -- 
> *************************************************
> Deborah MacPherson
> www.accuracyandaesthetics.com <http://www.accuracyandaesthetics.com>
> www.deborahmacpherson.com <http://www.deborahmacpherson.com>
> The content of this email may contain private
> confidential information. Do not forward, copy,
> share, or otherwise distribute without explicit
> written permission from all correspondents.
> **************************************************
>Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
>Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>    (012)

Patrick Durusau
Chair, V1 - Text Processing: Office and Publishing Systems Interface
Co-Editor, ISO 13250, Topic Maps -- Reference Model
Member, Text Encoding Initiative Board of Directors, 2003-2005    (013)

Topic Maps: Human, not artificial, intelligence at work!     (014)

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (015)

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