Re: University Website Developer is Fuzzy on AI

WSiler (wsiler@aol.com)
Mon, 17 Aug 1998 20:22:24 +0200 (MET DST)

>The next phase is to incorperate software capable of
>determining which general university requirements have been satisified,
>and to allows the rules and data to be eaisly changed by non-programmers
>from any web browser.

There are a couple of ways of doing this.

One way is to write your rules so as to interpret a data base. We read in the
data base, creating instances of a data structure in the expert system. We
write our rule(s) so that the antecedent uses the the data base, rather than
having fixed values built into the rule. Each instance of the data structure
can cause an instance of the rule(s) to be fireable. In this way, we can modify
the data base rather than the rules.

Some expert systems (OPS5, FLOPS) permit having a rule definition in the rule
consequent. This is a simple such construct:

Rule IF (in Data u =< A> and v = <B> and w = <C> and x = <D>...)
Then
rule IF (<A> = <B>) THEN <C> = <D>;

We have a data base which stores a set of values for x, y and z. These values
are read in as values for x, y and z in instances of a data structure Data. For
each such instance of data, a different instance of the single rule above will
be fireable. When these rule instances are fired (prefereably in parallel, as
most fuzzy expert systems operate) a different rule of the form given above
will be created. We have used this format to create rules for preliminary
classifications from a data base; if we wish to add, subtract or modify these
rules, we simply change the data base. The original program remains unaltered.

The basic thought is this: in an expert system you can store expert knowledge
either in rules or in data bases, whichever is better for your purpose. In
general, storing expert knowledge in data bases reduces the number of rules you
must write, provided that you can use different instances of a rule to do your
reasoning job, or if you need different rules of the same format (same number
of antecedent and consequent clauses of the same general type).

In either case, modifying a data base is a heck of a lot easier than modifying
rules.

William Siler


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