Spamprobe -D option

I'm currently using version 0.9f. Planning to upgrade to 1.4b shortly.

I've been using the -D option for a long time without really thinking about
it.  The shared db is being fed from a spamtrap address, so it never sees any
"good" mail.

I ran a couple experiments this morning to see what effect it is really
having.  I started by deleting the shared db and then training it with about
100 recent spams.  Then I took a collection of 60 spams of a specific type
that I'm interested in blocking better ("mexican pharmacy" spams) and scored
them using these 3 combinations:

1. -d shared
2. -d mine
3. -D shared -d mine

"shared" is the shared db, the training of which included a few of the mexican
pharmacy spams.  "mine" is my own db, which has not been trained with those
samples.  (Er, that's not true -- they've all been through "spamprobe train"
and marked as GOOD, and I have not corrected my db for these messages).

In case 1, it recognized all of the test spams as SPAM (score 1.0000).  In
case 2 it scored them all as GOOD (score 0.0000).  These were expected
results.

In case 3, it scored them all as GOOD (0.0000).  This was a little suprising.
I thought it would score them as SPAM, or at least at something higher than
0.0000.

If I train "mine" on just a few of the "mexican pharmacy" samples (I did 10 of
them), it starts scoring them all as SPAM, most at 1.0000, all over 0.9900.  I
seem to get the same scores for both case 2 and case 3.  i.e., the -D option
doesn't seem to affect the scoring.


I realize this testing isn't exhaustive.  I'm wondering if anyone else has had
similar or different experience with the -D option though.  Anyone?

Thanks,

Mark
-- 
Mark Costlow    | Southwest Cyberport | Fax:   +1-505-232-7975
[email protected] | Web:   www.swcp.com | Voice: +1-505-232-7992

abq-strange.com -- Interesting photos taken in Albuquerque, NM
                   Last post: Iris Door @ 2006-02-01 01:27:05


-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642