Re: Some questions setting up latest CRM114

   From: Leonard Lin <[email protected]>

   I have some more free time recently and have been playing around w/  
   the latest versions of CRM114 and had a couple questions: I know that  
   if you use the reaver_cache you can avoid this, but if the cache is  
   missing, does mailreaver.crm filter out X-CRM114 headers in messages  
   or do I need to do that if I'm doing my own training?

   i.e. I wrote a python script so that I could train complete sets via  
   IMAP which is fine with my own clean corpus (in IMAP folders), but  
   doesn't strip X-CRM114 headers at the moment: http://jabba.randomfoo.decenturl.com/crm114-trainer.py 
      -- is this something I can run (zapping everything now that I also  
   have procmailed items also being move in there?

   Also, I'm using the 'hyperspace unique' classifier (it seemed like it  
   had a lot less unsures early on - is it more accurate overall?) w/ the  
   default pR threshold of 10 (I'm using 20070810-BlameTheSegfault).   
   I've trained 276 Ham and 172 Spam (in date order, ToE and on UNSUREs)  
   - as of right now it looks like most new mail is still going in the  
   unsure folder still - is this normal?  At what point does it usually  
   level off?

It should be pretty good now.

I suggest looking at your unsures and seeing if there are definitive
pR values where

   - Everthing with a pR greater than this is GOOD.

   - Everything with a pR less than (minus this PR) is SPAM

and adjust your threshold to that value.

I'm finding that a pR threshold of +/- 5.0 is working better these
days, as it allows micro spams to be caught very easily (microspams 
are spams with just a line of text and single URL as the payload)

    - Bill Yerazunis


-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace