Newsgroups: comp.ai.shells,comp.ai
Path: cantaloupe.srv.cs.cmu.edu!rochester!udel!gatech2!swrinde!newsfeed.internetmci.com!in2.uu.net!tron!usenet
From: kneale.j.b@wec.com (Jim Kneale)
Subject: Re: ATTAR's XpertRule
Sender: usenet@tron.bwi.wec.com (USEnet News Poster)
Message-ID: <DI7Ito.7uC@tron.bwi.wec.com>
Date: Fri, 17 Nov 1995 21:39:24 GMT
X-Nntp-Posting-Host: sse1pc92.bwi.wec.com
References: <DI575u.9pE@info.uucp>
Mime-Version: 1.0
Organization: Westinghouse Electric Corp.
X-Newsreader: WinVN 0.99.2
Lines: 29
Xref: glinda.oz.cs.cmu.edu comp.ai.shells:2844 comp.ai:34810

In article <DI575u.9pE@info.uucp>, sout6715@mach1.wlu.ca says...
>
>Is anyone out there using this product? How does it stack up against 
>forward/backward chaining tools, such as Aion's inference engine?

The latest issue of PCAI magazine carries a review of XpertRule.  We 
purchased it, and it appears to be extremely powerful if tree-based 
reasoning will work in your app.  I didn't get real far into it, but it 
appears to want you to follow its method for running the consultation.  
If you have a lot of pre-existing facts that will almost complete a 
branch of the tree, it won't go after it until it has gotten around to 
it.  _Some_ forward/back chaining systems will look at which rules are 
closest to firing and ask for the facts on those first.

Also, be advised that unless you are using it in the USA and register it 
with your name, it must be used with a key.  In addition, you must store 
it on the hard drive of the PC that is executing it.  File server 
operation, even with a key, is prohibited by the license.  We can't use 
the license-monitoring software on our file servers with it like other 
programs, and we can't protect it from getting erased off the C: drive by 
the users.

-- 
Jim Kneale               kneale.j.b@wec.com
ph  (410)712-2248   ///  Westinghouse Electric Corp
fax (410)712-2815  (..)  Baltimore MD USA 
----------------ooO-()-Ooo------------------
#include "std_disclaimer.h" //opinions are my own

