CG: Modèle, formalisme, langage

[lynda souadih]

Bonjour,

Je trouve dans la littérature plusieurs définitions des CG, je trouve
que les GC sont un formalisme, un modèle et un langage.

Je veut savoir c'est quoi au juste les GC conceptuel, un modèle? un
formalisme? ou un langage? je suppose que les trois définitions sont
justes, mais je veux une petite explication si c'est possible.

Merci beaucoup beaucoup à vous.

---------------------------------------------------------------------
To unsubscribe, e-mail: cg-unsubscribe@...
For additional commands, e-mail: cg-help@...

[John F. Sowa]

Lynda,

 > Je veut savoir c'est quoi au juste les GC conceptuel,
 > un modèle? un formalisme? ou un langage?

I would say that conceptual graphs are a mathematical formalism
(which is how they were defined in Chapter 3 of my 1984 book)

  1. that can be used as a dialect of ISO standard Common Logic,

  2. that can be used as a graphic language for people to see,

  3. that can be used as a linear language (CGIF) for computers
     to transmit among themselves.

Following is a short summary:

    http://www.jfsowa.com/cg/cg_hbook.pdf

John Sowa


---------------------------------------------------------------------
To unsubscribe, e-mail: cg-unsubscribe@...
For additional commands, e-mail: cg-help@...

[lynda souadih]

Thank you very much for the answer. What I do not understand it is the
difference between model, formalism, and language.

2009/9/28, John F. Sowa <sowa@...>:

> Lynda,
>
>  > Je veut savoir c'est quoi au juste les GC conceptuel,
>  > un modčle? un formalisme? ou un langage?
>
> I would say that conceptual graphs are a mathematical formalism
> (which is how they were defined in Chapter 3 of my 1984 book)
>
>   1. that can be used as a dialect of ISO standard Common Logic,
>
>   2. that can be used as a graphic language for people to see,
>
>   3. that can be used as a linear language (CGIF) for computers
>      to transmit among themselves.
>
> Following is a short summary:
>
>     http://www.jfsowa.com/cg/cg_hbook.pdf
>
> John Sowa
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: cg-unsubscribe@...
> For additional commands, e-mail: cg-help@...
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: cg-unsubscribe@...
For additional commands, e-mail: cg-help@...

[John F. Sowa]

Lynda,

Those words have been used in many different ways:

 > What I do not understand is the difference between
 > model, formalism, and language.

The word 'model' is the most confusing.  In logic, there is
a clear definition:  A model for a set of axioms is a set of
entities and a set of relations among those entities for which
the axioms are true.

But in the 1970s, the term 'data model' was introduced to
distinguish relational databases from network databases.
People would talk about the 'relational model', which was
usually displayed as a collection of tables, vs. the
'network model', which was usually displayed as a large
graph structure.

The distinction defined by the term 'data model' has no basis
in logic, because every set of relationships that can be expressed
in a relational database can be translated to a logically equivalent
form in a network database and vice versa.

Since the two kinds of databases are isomorphic in a mathematical
sense, it is extremely confusing to use the word 'model' as a
term for distinguishing them.  It is even more confusing because
exactly the same languages (such as SQL or Datalog) can be used
to define them or to access them.

There is, of course, a difference in the relative efficiency of
certain kinds of operations that may be performed on a network
or on a collection of tables.  That difference may be critical
to a system programmer who has to optimize performance, but it is
a low-level optimization that should be invisible to most users.

A formalism is usually something that is defined by mathematical
axioms.  In that sense CGs are a formalism because they were
defined axiomatically from the very beginning.  However, they
can also be used in a very informal way, and that is one of
their advantages:  they can be used in a systematic methodology
for a step-by-step translation from a vague, informal statement
to a more formal one.  Following are some slides that discuss
that point:

    http://www.jfsowa.com/talks/cmapping.pdf

And finally a language is usually something that is defined
by a grammar.  The linear notation for CGIF is defined by a
traditional grammar, but it is also possible to use graph
grammars to define CGs as graphs.  The canonical formation
rules are, in fact, a kind of graph grammar that can be used
to derive new CGs from a starting set of CGs.  See Section 3
of the following paper for a brief summary of the canonical
formation rules and their relationship to the rules of
inference of logic:

    http://www.jfsowa.com/cg/cg_hbook.pdf

John Sowa










---------------------------------------------------------------------
To unsubscribe, e-mail: cg-unsubscribe@...
For additional commands, e-mail: cg-help@...