Q: negative fuzzy facts & fuzzy fact with a CF

Cate ten BD (bdtcate@kits.cs.vu.nl)
Mon, 23 Aug 1999 15:25:04 +0200 (MET DST)

Hello everybody,

I have two problems implementing a small knowledge system in
FuzzyCLIPS. For both problems I have a solution, only I don't
know if they are the best ones (maybe I oversee some traps)
and I really need some literature references to justify the
measures taken. I would be very thankful if anybody could
supply me with notes, alternatives and/or references.

I. Conjunction instead of disjunction when asserting negative
fuzzy facts.

If three rules conclude the following statements about the
temperature

r1: "the temperature is low"
r2: "the temperature high"
r3: "the temperature is not very high"

then I want those statement to be combined as "the temperature is
(low or high) and not very high". So I make a distinction between
positive statements p_1..p_n and negative statements n_1..n_n, and
then I combine them as (p_1 OR .. OR p_n) AND (n_1 OR .. OR n_n).

II. Asserting fuzzy facts with a certainty factor.

If two rules conclude about the temperature:

r1: "the temperature is low, with certainty 0.3"
r2: "the temperature is high, with certainty 0.8"

Then I want something like:

+------------------------+
| _____|
| _ - |
|---- _ _- |
| - - |
| - _ - |
+------------------------+

which I did by multiplying each curve before taking the union.
However this does not give a full account of things. For example
two rules concluding the same linguistic expression with the same
certainty should combine to give a higher certainty factor
conclusion. I think that it's enough to solve the problems with
implementing the knowledge though.

Thanks, Balder.

############################################################################
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