                                                                                                                                                                                                                                                                
Date: Wed, 19 Sep 2012 07:04:32 -0700 (PDT)
From: Matteo Dell'Amico <matteodellam...@gmail.com>
To: pyke@googlegroups.com
Message-Id: <4a9a665b-1a89-474a-bc7d-a1cba2a5813e@googlegroups.com>
Subject: Assert universal fact in FC rules?
MIME-Version: 1.0
Content-Type: multipart/mixed; 
	boundary="----=_Part_42_1353342.1348063472520"

------=_Part_42_1353342.1348063472520
Content-Type: multipart/alternative; 
	boundary="----=_Part_43_29271166.1348063472520"

------=_Part_43_29271166.1348063472520
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit

Hello,

thanks for Pyke, it's really a nice piece of work! :)

I am considering using forward chaining to perform "pre-computation" of 
facts that you can infer from a .kfb file. Say, for the standard family 
relations example, I'd like to precompute an "is_ancestor" fact for all 
such relations, on the basis that I'm going to need it often and it would 
be a waste to recompute it with backward chaining any time it is needed. 
The problem is, when I reset the knowledge engine, as far as I understand 
facts generated through forward chaining are all erased because they are 
specific facts and not universal fact.

So, I'd need a way to either a) save specific facts in a new fact base or 
b) directly assert universal facts from forward chaining.

Would that be possible in Pyke?

thanks! :)
matteo

------=_Part_43_29271166.1348063472520
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 7bit

Hello,<br><br>thanks for Pyke, it's really a nice piece of work! :)<br><br>I am considering using forward chaining to perform "pre-computation" of facts that you can infer from a .kfb file. Say, for the standard family relations example, I'd like to precompute an "is_ancestor" fact for all such relations, on the basis that I'm going to need it often and it would be a waste to recompute it with backward chaining any time it is needed. The problem is, when I reset the knowledge engine, as far as I understand facts generated through forward chaining are all erased because they are specific facts and not universal fact.<br><br>So, I'd need a way to either a) save specific facts in a new fact base or b) directly assert universal facts from forward chaining.<br><br>Would that be possible in Pyke?<br><br>thanks! :)<br>matteo<br>
------=_Part_43_29271166.1348063472520--

------=_Part_42_1353342.1348063472520--