[Top] [All Lists]

Re: [ontolog-forum] CL, CG, IKL and the relationship between symbols in

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Tue, 08 Jan 2008 14:03:35 -0500
Message-id: <4783C907.1000307@xxxxxxxxxxx>
Duane and Peter,    (01)

The number of possible qualifiers is open ended -- potentially
infinite.  The *worst* possible thing anyone could ever do is
to design a computer notation to support every one of those
options.  Instead, I recommend that you write down exactly what
you mean in your native language.  Then study what you have
written in NL before you even begin to worry about how to
represent the results in *any* formal notation.    (02)

DN> As the BuyerParty vs. Seller Party hierarchic context qualifies
 > the data element.
 > Is this the type of syntax context you refer to?    (03)

PFB> ... in Iceland, for example, the first name is the significant
 > name - even the telephone directories are ordered by first
 > name; the family name is simply a sort of qualifier (daughter of...,
 > son of ...) but which is not considered as significant as the first
 > name...
 > ... the problem is, however careful you are, there will often be
 > situations where the context requires something else.    (04)

When translating any kind of knowledge to any formalism (i.e., any
version of logic or any computer representation including Common Logic,
the Semantic Web notations, etc.) you should ignore the syntax of the
formalism and say exactly what you mean in very simple sentences in
your own native language.    (05)

In this case, start by listing the sentences you want to express:    (06)

  1. Every person has a name.    (07)

  2. The name has several parts.    (08)

  3. In some countries, the two main parts of a person's name
     are the family name and the given name.    (09)

  4. In Iceland, the two main parts of a person's name are
     the given name and the patronymic.    (010)

  5. A person's patronymic is derived from the given name
     of the father of that person.    (011)

  6. A person may play many different roles.    (012)

  7. A person who buys something plays the role of buyer.    (013)

  8. A person who sells something plays the role of seller.    (014)

These eight sentences are independent of any formalism for logic
or any computer system.  But once you have listed them in this
form, you have the basis for translating them to any ontology
stated in any kind of logic, UML diagrams, database language,
Semantic Web notation, etc., etc., etc.    (015)

Note several different points:    (016)

  1. The central focus is the human being or person.    (017)

  2. The next most important point is the name.    (018)

  3. That name be represented in different ways.    (019)

  4. That person can play different roles.    (020)

  5. The roles are directly related to the person,
     indirectly related to the person's name, and
     very indirectly related to the parts of the name
     in different representations.    (021)

Every one of the eight sentences above plus these five points
can be translated from English directly to Common Logic.    (022)

In fact, we have a particular English-like notation, called
Common Logic Controlled English (CLCE), which looks very much
like the kinds of English sentences stated above.  And CLCE
can be automatically translated to Common Logic.    (023)

That is my primary recommendation for how to do ontology:
use a controlled NL as the *primary* notation.  Use various
computer languages, such as CL or UML or RDF or OWL as
*secondary* languages used for a particular implementation.    (024)

Fundamental principle:  If you want somebody to understand
ontology, don't *ever* teach them RDF, OWL, or any other
computer notation until *after* they are capable of stating
exactly what they mean in their own native language.  Otherwise,
their ideas will be forever polluted by the quirks of that
notation.    (025)

John    (026)

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    (027)

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