[ont.events] U of Toronto Computer Science colloquium, Feb. 10

clarke@utcsri.UUCP (02/03/87)

COLLOQUIUM, Tuesday, February 10, 11 am, SF1101
         (SF = Sandford Fleming Building, 10 King's College Road)

                           Professor L.A. Zadeh
                         University of California

            ``The Role of Fuzzy Logic in Common Sense Reasoning
                       and Knowledge Representation"


     Among the best-known approaches to commonsense reasoning are default
reasoning and non-monotonic logic.  In an alternative approach which is
described in our talk, the conceptual framework of fuzzy logic is employed
as a basis for both representation of and inference from commonsense
knowledge.  The point of departure in our approach is the assumption that
commonsense knowledge may be represented as a collection of dispositions,
that is, propositions which are preponderantly, but no necessarily always,
true.  Through a process of explicitation, dispositions may be transformed
into dispositional propositions with explicit fuzzy quantifiers, e.g.,
most, almost all, usually, etc.  Then, a variety of fuzzy syllogisms may be
employed to infer a dispositional conclusion from dispositional premises.
And finally, the dispositional conclusion may be transformed into a dispo-
sition through the suppression of fuzzy quantifiers.  Viewed in this per-
spective, non-monotonic reasoning may be seen as a monotonic reasoning with
dispositions.

     The concept of a disposition gives rise to a number of other disposi-
tional concepts, among them dispositional predicates, dispositional com-
mands, dispositional relations, etc.  Possible applications of disposi-
tional concepts are described and illustrated by examples.
-- 

Jim Clarke -- Dept. of Computer Science, Univ. of Toronto, Canada M5S 1A4
              (416) 978-4058
{allegra,cornell,decvax,linus,utzoo}!utcsri!clarke