[ACT-R-users] ACT-R robot (UNCLASSIFIED)

[Kelley, Troy (Civ,ARL/HRED)]

Classification:  UNCLASSIFIED 
Caveats: NONE

Susan,

	We have been looking at using a semantic network as a starting point
for declarative memory chunks.  There are two major semantic networks that
we are looking at using.  ConceptNet from MIT and OpenCYC which is a spin
off of Doug Lenat's work at the University of Texas.  We are still working
on exactly how to interface a semantic network with declarative memory
chunks, but I think the general idea has promise. We are also trying to use
a production syntax so that the same general productions don't have to be
developed over and over.  Both of these ideas have promise but we are still
working on the implementations.

Troy D. Kelley
AMSRD-HR-SE
Army Research Laboratory
Human Research and Engineering Directorate (HRED)
Aberdeen Proving Ground, Aberdeen MD 21005-5425
voice: 410-278-5869
fax: 410-278-9523

-----Original Message-----
From: act-r-users-bounces at act-r.psy.cmu.edu
[mailto:act-r-users-bounces at act-r.psy.cmu.edu] On Behalf Of Susan Chipman
Sent: Tuesday, June 24, 2008 1:10 PM
To: Dan Bothell
Cc: act-r-users at act-r.psy.cmu.edu
Subject: Re: [ACT-R-users] ACT-R robot

        As the former sponsor of much of the ACT-R work, I would really like
to have seen a cumulative library of model building work in order to avoid
duplication of effort and the associated waste of scarce research funding.
In particular, it seemed to me that declarative knowledge should not be
built over and over again.   I retired, so I am no longer in a position to
push this agenda, but I hope the community will come to such a conclusion
soon.
 
Susan Chipman


On Tue, Jun 24, 2008 at 12:00 PM, Dan Bothell <db30 at andrew.cmu.edu> wrote:




	--On Tuesday, June 24, 2008 8:04 AM -0700 Bruce J Weimer MD
	
	<bjweimer at charter.net> wrote:
	
	
	> Also, I was wondering as I read the papers, is information that's
learned
	> during a session with the robot stored or saved when the robot's
turned
	> off  or is it lost?  I'm reading through the ACT-R tutorials and
it seems
	> that  each time a model is run, it's reset to run again - there
(so far)
	> hasn't  been any mention of a mechanism to save the information in
	> declarative  memory, weights, latencies, etc.
	>
	
	
	There is no mechanism built into ACT-R for saving the internal state
of any
	module or the system in general.  However, depending on what you
want to
	save, it's usually not too difficult to just cache that out and read
it
	back in later.
	
	For example, here's a function which will create a file with the
commands
	to rebuild the chunks in declarative memory and restore their
parameters
	(only one of the declarative parameters is included here for
brevity):
	
	(defun save-chunks (file-name)
	 (let ((chunks (no-output (dm)))
	       (params (no-output (sdp :name :reference-count)))
	       (cmdt (car (no-output (sgp :cmdt)))))
	
	   (sgp-fct (list :cmdt file-name))
	
	   (command-output "(add-dm ")
	
	   (dolist (x chunks)
	     (command-output "(")
	     (pprint-chunks-fct (list x))
	     (command-output ")"))
	   (command-output ")")
	
	   (dolist (x params)
	     (command-output "(sdp ~A :reference-count ~s)" (first x)
(second x)))
	
	   (sgp-fct (list :cmdt cmdt))))
	
	The file that writes out could then just be loaded later to restore
the
	model's declarative memory.  A similar function could be written to
dump
	the productions and their parameters (see the reference manual for
the
	relevant commands).
	
	Depending on how the general parameters for the model are set there
may
	be some other things that would need to be adjusted as well, but
those
	sorts of things are going to vary from model to model.  One example
would
	be either advancing the time to where the "old" model left off or
adjusting
	all of the declarative parameters to reflect the "current" time
being
	necessary if base-level learning were enabled.
	
	Dan
	
	_______________________________________________
	ACT-R-users mailing list
	ACT-R-users at act-r.psy.cmu.edu
	http://act-r.psy.cmu.edu/mailman/listinfo/act-r-users
	


Classification:  UNCLASSIFIED 
Caveats: NONE

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 6040 bytes
Desc: not available
URL: <http://mailman.srv.cs.cmu.edu/pipermail/act-r-users/attachments/20080624/41034a62/attachment.bin>