abstract, talk


Subject: abstract, talk
From: Michelle T. Lin (michlin@cs.berkeley.edu)
Date: Mon Jan 31 2000 - 20:58:37 MET


*********************************************************************
Berkeley Initiative in Soft Computing (BISC)
*********************************************************************

NOTE: For technical reasons, this announcement was not sent on time.
Its purpose at this point is to provide information about the talk
that took place already.

-------------

Please note that the BISC Seminar lecture on "Toward an Enlargement of
Role of Natural Languages in Information Processing, Decision and
Control" by Lotfi A. Zadeh, is scheduled for Tuesday, January 25,
2000, in Room 405 Soda Hall, 4-5 pm.

Subsequent meetings by the Seminar will be held on Tuesdays, 4-5 pm,
in Room 310 Soda Hall.

-----------
Toward an Enlargement of the Role of Natural Languages in
Information Processing, Decision and Control

Lotfi A. Zadeh*

Abstract

        It is a deep-seated tradition in science to view the use of
natural languages in scientific theories as a manifestation of
mathematical immaturity. The rationale for this tradition is that
natural languages are lacking in precision. However, what is not
recognized to the extent that it should, is that adherence to this
tradition carries a steep price. In particular, a direct consequence
is that existing scientific theories do not have the capability to
operate on perception-based information exemplified by "Most Finns are
honest." Such information is usually described in a natural language
and is intrinsically imprecise, reflecting a fundamental limitation on
the cognitive ability of humans to resolve detail and store
information. Because of their imprecision, perceptions do not lend
themselves to meaning-representation through the use of precise
methods based on predicate logic. This is the principal reason why
existing scientific theories do not have the capability to operate on
perception-based information.

        In a related way, the restricted expressive power of
predicate-logic-based languages rules out the possibility of defining
many basic concepts such as causality, resemblance, smoothness and
relevance in realistic terms. In this instance, as in many others,
the price of precision is over-idealization and lack of robustness.

        In a significant departure from existing methods, in the
approach which is described in this talk the high expressive power of
natural languages is harnessed by constructing what is called a
precisiated natural language (PNL). In essence, PNL is a subset of a
natural language (NL) -- a subset which is equipped with
constraint-centered semantics (CSNL) and is translatable into what is
called the Generalized Constraint Language (GCL). A concept which has
a position of centrality in GCL is that of a generalized constraint
expressed as X isr R, where X is the constrained variable, R is the
constraining relation, and isr (pronounced as ezar) is a variable
copula in which r is a discrete-valued variable whose value defines
the way in which R constrains X. Among the principal types of
constraints are possibilistic, veristic, probabilistic, random-set,
usuality, and fuzzy-graph constraints.

        With these constraints serving as basic building blocks, more
complex (composite) constraints may be constructed through the use of
a grammar. The collection of composite constraints forms the
Generalized Constraint Language (GCL). The semantics of GCL is
defined by the rules that govern combination and propagation of
generalized constraints. These rules coincide with the rules of
inference in fuzzy logic (FL).

        A key idea in PNL is that the meaning of a proposition, p, in
PNL may be represented as a generalized constraint which is an element
of GCL. Thus, translation of p into GCL is viewed as an explicitation
of X, R and r. In this sense, translation is equivalent to
explicitation.

        The concept of a precisiated natural language and the
associated methodologies of computing with words and the computational
theory of perceptions open the door to a wide-ranging generalization
and restructuring of existing theories, especially in the realms of
information processing, decision and control. In this perspective,
what is very likely is that in coming years a number of basic concepts
and techniques drawn from linguistics will be playing a much more
important role in scientific theories than they do today.

---------
*Professor in the Graduate School and Director, Berkeley
Initiative in Soft Computing (BISC), Computer Science Division and the
Electronics Research Laboratory, Department of EECS, University of
California, Berkeley, CA 94720-1776; Telephone: 510-642-4959; Fax:
510-642-1712; E-Mail: zadeh@cs.berkeley.edu. Research supported in
part by NASA Grant NAC2-1177, ONR Grant N00014-96-1-0556, ONR Grant
FDN0014991035, ARO Grant DAAH 04-961-0341 and the BISC Program of UC
Berkeley.

---------------------------------

Toward an Enlargement of the Role of Natural Languages in
Information Processing, Decision and Control

Lotfi A. Zadeh*

Extended Abstract

        It is a deep-seated tradition in science to view the use of
natural languages in scientific theories as a manifestation of
mathematical immaturity. The rationale for this tradition is that
natural languages are lacking in precision. However, what is not
recognized to the extent that it should, is that adherence to this
tradition carries a steep price. In particular, a direct consequence
is that existing scientific theories do not have the capability to
operate on perception-based information exemplified by "Most Finns are
honest." Such information is usually described in a natural language
and is intrinsically imprecise, reflecting a fundamental limitation on
the cognitive ability of humans to resolve detail and store
information. Because of their imprecision, perceptions do not lend
themselves to meaning-representation through the use of precise
methods based on predicate logic. This is the principal reason why
existing scientific theories do not have the capability to operate on
perception-based information.

        In a related way, the restricted expressive power of
predicate-logic-based languages rules out the possibility of defining
many basic concepts such as causality, resemblance, smoothness and
relevance in realistic terms. In this instance, as in many others,
the price of precision is over-idealization and lack of robustness.

        In a significant departure from existing methods, in the
approach which is described in this talk the high expressive power of
natural languages is harnessed by constructing what is called a
precisiated natural language (PNL). In essence, PNL is a subset of a
natural language (NL) -- a subset which is equipped with
constraint-centered semantics (CSNL) and is translatable into what is
called the Generalized Constraint Language (GCL). A concept which has
a position of centrality in GCL is that of a generalized constraint
expressed as X isr R, where X is the constrained variable, R is the
constraining relation, and isr (pronounced as ezar) is a variable
copula in which r is a discrete-valued variable whose value defines
the way in which R constrains X. Among the principal types of
constraints are the following: possibilistic constraint, r=blank,
with R playing the role of the possibility distribution of X; veristic
constraint, r=v, in which case R is the verity (truth) distribution of
X; probabilistic constraint, r=p, in which case X is a random variable
and R is its probability distribution; r=rs, in which case X is a
fuzzy-set-valued random variable (fuzzy random set) and R is its
fuzzy-set-valued probability distribution; and fuzzy-graph constraint,
r=fg, in which case X is a fuzzy-set-valued variable and R is its
fuzzy-set-valued possibility distribution.

        With these constraints serving as basic building blocks, which
are analogous to terminal symbols in a formal language, more complex
(composite) constraints may be constructed through the use of a
grammar. Simple examples of composite constraints are: X isr R and X
iss S; and, if X isr R then Y iss S, or, equivalently, Y iss S if X
isr R. The collection of composite constraints forms the Generalized
Constraint Language (GCL). The semantics of GCL is defined by the
rules that govern combination and propagation of generalized
constraints. These rules coincide with the rules of inference in
fuzzy logic (FL).

        A key idea in PNL is that the meaning of a proposition, p, in
PNL may be represented as a generalized constraint which is an element
of GCL. Thus, translation of p into GCL is viewed as an explicitation
of X, R and r. In this sense, translation is equivalent to
explicitation. The translate of p in GCL is referred to as its
canonical form (CF(p)). These concepts form the core of what is
called constraint-oriented semantics of natural languages (CSNL).

        In what ways does the concept of a precisiated natural
language serve to enlarge the role of natural languages in scientific
theories? Standing out in importance are the following.

        a. By serving as a base for the methodologies of computing
with words (CW) and the computational theory of perceptions (CTP).
The importance of these methodologies derives from the fact that in
real-world settings decision-relevant information is frequently a
mixture of measurements and perceptions. Existing methods of decision
analysis are measurement-based and do not have the capability to
operate on perception-based information. The computational theory of
perceptions adds this capability to existing theories and thereby
enhances their ability to deal with realistic problems.

        b. By making it possible to compute with words rather than
with numbers when numerical values of decision-relevant variables are
not known with sufficient precision to justify the use of
numerically-based methods.

        c. By making it possible to define concepts that are
perception - rather than measurement-based. Examples of such concepts
are: the concept of a usual value of a random variable; the concept
of smoothness of a function; the concept of optimality in
multicriterion optimization; and the concept of relevance in database
search. In particular, in the context of decision analysis, it
becomes possible and, in many cases, advantageous, to define the
goals, constraints, dependencies, probabilities, utilities, and
risk-aversion in PNL rather than analytically or numerically. These
are but a few examples of ways in which PNL-based methods can be used
as addenda to existing numerically-based methods.

        The concept of a precisiated natural language and the
associated methodologies of computing with words and the computational
theory of perceptions open the door to a wide-ranging generalization
and restructuring of existing theories, especially in the realms of
information processing, decision and control. In this perspective,
what is very likely is that in coming years a number of basic concepts
and techniques drawn from linguistics will be playing a much more
important role in scientific theories than they do today.

-------
*Professor in the Graduate School and Director, Berkeley
Initiative in Soft Computing (BISC), Computer Science Division and the
Electronics Research Laboratory, Department of EECS, University of
California, Berkeley, CA 94720-1776; Telephone: 510-642-4959; Fax:
510-642-1712; E-Mail: zadeh@cs.berkeley.edu. Research supported in
part by NASA Grant NAC2-1177, ONR Grant N00014-96-1-0556, ONR Grant
FDN0014991035, ARO Grant DAAH 04-961-0341 and the BISC Program of UC
Berkeley.

----------
If you ever want to remove yourself from this mailing list,
you can send mail to <Majordomo@EECS.Berkeley.EDU> with
the following command in the body of your email message:
unsubscribe bisc-group
or from another account,
unsubscribe bisc-group <your_email_adress>

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Do NOT send unsubscribe requests to bisc-group@cs.berkeley.edu
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
If you ever want to remove yourself from this mailing list,
you can send mail to <Majordomo@EECS.Berkeley.EDUwith the following
command in the body of your email message:
unsubscribe bisc-group
or from another account,
unsubscribe bisc-group <your_email_adress>
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Do NOT send unsubscribe requests to bisc-group@cs.berkeley.edu
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

############################################################################
This message was posted through the fuzzy mailing list.
(1) To subscribe to this mailing list, send a message body of
"SUB FUZZY-MAIL myFirstName mySurname" to listproc@dbai.tuwien.ac.at
(2) To unsubscribe from this mailing list, send a message body of
"UNSUB FUZZY-MAIL" or "UNSUB FUZZY-MAIL yoursubscription@email.address.com"
to listproc@dbai.tuwien.ac.at
(3) To reach the human who maintains the list, send mail to
fuzzy-owner@dbai.tuwien.ac.at
(4) WWW access and other information on Fuzzy Sets and Logic see
http://www.dbai.tuwien.ac.at/ftp/mlowner/fuzzy-mail.info
(5) WWW archive: http://www.dbai.tuwien.ac.at/marchives/fuzzy-mail/index.html



This archive was generated by hypermail 2b25 : Thu Apr 06 2000 - 15:59:41 MET DST