Xref: utzoo comp.ai:2322 comp.cog-eng:650
Path: utzoo!attcan!uunet!husc6!uwvax!rutgers!ucsd!ucsdhub!hp-sdd!ncr-sd!serene!pnet12!bstev
From: bstev@pnet12.cts.com (Barry Stevens)
Newsgroups: comp.ai,comp.cog-eng
Subject: Re: CFP: 2nd Generation Expert Systems
Message-ID: <258@serene.CTS.COM>
Date: 1 Oct 88 14:56:13 GMT
Sender: root@serene.CTS.COM
Organization: People-Net [pnet12], Del Mar, Ca.
Lines: 23

jbn@glacier.STANFORD.EDU (John B. Nagle) writes:
>
>      Do any working and useful "second generation expert systems" actually
>exist?  Or are these guys just looking for new suckers to pour money in?
>
>					John Nagle

I have experimented with a generalized model of a large datacenter as a
driver for a diagnostic expert system, to answer questions such as "Why did
my job fail?" "Why is response time slow?" "Who has to be told that I'm
going to change program xyz?" "Why are we over budget?" "Where are there
opportunities to decrease operating costs?"

The rules have been constructed, some testing done. Waiting for me to have
time to deal with the complexities of asking english language questions
instead of selecting from menus. Working? Sort of. Useful? Not yet.

Since I'm not schooled in the scientific aspects of these things, I am not
sure whether this sort of thing qualifies as a part of this discussion.

UUCP: {crash ncr-sd}!pnet12!bstev
ARPA: crash!pnet12!bstev@nosc.mil
INET: bstev@pnet12.cts.com