garbage removal and 'outsiders noise'

[Jim Correia]

On Thursday, April 17, 2003, at 07:43  PM, Greg Louis wrote:

> To me, 10% fn would be a big number.

Well I'd like it to be lower, but that is currently what it is (no fp 
though.)

> I don't run with -u, but train manually: copy all mail to a single mbox
> file, and periodically use bogofilter to break it in 3: spam, nonspam,
> unsure.

Is there a reason you do it this way?

Theoretically, since it "learns" by seeing representative spam and 
non-spam messages, shouldn't more data produce more accurate results, 
all other things equal?

(I know you previously commented about the lopsidedness of my word 
lists...)

Jim