(no subject)

To: Thomas Michael Hagen <[email protected]>
CC: [email protected]
In-reply-to: <d77c32370905060810r3099c93je57ded9dd5b791f2-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
	(message from Thomas Michael Hagen on Wed, 6 May 2009 17:10:57 +0200)
Subject: Re: [Crm114-general] ready to roll version of newspaper article topic
	classifier
References:  <d77c32370905060810r3099c93je57ded9dd5b791f2-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>


   From: Thomas Michael Hagen <[email protected]>

   i'm on my fifth version now, after a lot of work, and a lot of
   valuable input from this list.

   eight people in my department are now going to train the classifier by
   categorizing about a thousand articles each.

   here is the core of the program they are using:

   i hope this can be of use to some, and i always love harsh feedback :)


OK.... I think you might have an infinite loop.  Scroll on down....

        - Bill

   :try-to-classify: (:article:)
   {
       #THE CLASSIFICATION HALF:
       {
	   {
	       isolate (:stats:)
	       classify <osbf microgroom> [:article:] \
		       (bus.cfc cli.cfc con.cfc cul.cfc dom.cfc \
		       eco.cfc edu.cfc ene.cfc ent.cfc env.cfc \
		       for.cfc hea.cfc loc.cfc pol.cfc sci.cfc \
		       spo.cfc tec.cfc tra.cfc wor.cfc | unc.cfc out.cfc) (:stats:)
	       match (:: :class-conc:) [:stats:] /(...)\.cfc/ #perhaps pick the
   top 3 cats, and tag article with all 3 (if above some threshold)
	       { #classify succeeded, checking pR:
		   match (:: :pR:) [:stats:] /Best match.*: (.*)  \nTotal/
		   output /pR::*:pR:. /
		   eval /:@::*:pR: > 30:/ #play around with this number
		   output /Classify succeeded as ':*:class-conc:'\n/
		   return
	       }
	       output /Classification NOT confident! Please classify:\n\n/
	   }
	   alius
	   {
	       output /Classify failed! Please classify:\n\n/
	   }
       }
       #compare a low conf answer with the result of try-to-learn here,
   and use it if they are the same
       isolate (:conc-file:) /:*:class-conc:.cfc/
       isolate (:interaction:) /:*:class-conc::*:article::*:stats:/
       call /:interact:/ [:*:interaction:] #asking the user for the right topic
       { #if the answer was right, use that answer, and refute fram all
   other categories:
	   match [:user-feedback:] /^y$/
	   isolate (:user-feedback:) /:*:class-conc:/
	   isolate (:refute-args:) /:*:conc-file::*:article:/
	   call /:refute:/ [:*:refute-args:]
       }
       #THE LEARNING HALF:
       isolate (:user-feedback-conc-file:) /:*:user-feedback:.cfc/
       isolate (:count:) /7/
       isolate (:print-p:) /1/
       { # entering the training loop
	   learn <osbf microgroom> (:*:user-feedback-conc-file:) [:article:]
	   {
	       match [:print-p:] /1/
	       output /\nLearning as :*:user-feedback-conc-file:...\n/
	       isolate (:print-p:) /0/
	   }
	   isolate (:stats2:)
	   { #reclassifying to check improvement:
	       classify <osbf microgroom> [:article:]	\
		       (bus.cfc cli.cfc con.cfc cul.cfc dom.cfc \
		       eco.cfc edu.cfc ene.cfc ent.cfc env.cfc \
		       for.cfc hea.cfc loc.cfc pol.cfc sci.cfc \
		       spo.cfc tec.cfc tra.cfc wor.cfc | unc.cfc out.cfc) (:stats2:)
	       isolate (:pR2:) #here, you assume it ends up with the same cat as
   before. what if it doesn't?
	       match (:: :pR2:) [:stats2:] /Best match.*: (.*)  \nTotal/
	       isolate (:improvement:)
	       eval (:improvement:) /:@:(:*:pR2: - :*:pR:):/
	       output /imp::*:improvement:\n/
	       eval /:@::*:improvement: > 3:/
	       eval /:@::*:pR2: > 10:/ #play around with this number too
	       output /Succesfully learned! Moving on\n/
	       return
	   }
	   eval (:count:) /:@::*:count: - 1:/
	   {
	       eval /:@::*:count: > 0:/
	       liaf
	   }

Right there ^^^^^^   if :count: is > 0, the EVAL passes (and falls through),
to the LIAF, which sends it back up to the opening "{", then the EVAL 
passes, it falls through, whcih sends it up to the opening "{",  and so
on.

Or do I have it wrong?


------------------------------------------------------------------------------
The NEW KODAK i700 Series Scanners deliver under ANY circumstances! Your
production scanning environment may not be a perfect world - but thanks to
Kodak, there's a perfect scanner to get the job done! With the NEW KODAK i700
Series Scanner you'll get full speed at 300 dpi even with all image 
processing features enabled. http://p.sf.net/sfu/kodak-com