unsure results
David Relson
relson at osagesoftware.com
Tue Mar 2 13:20:27 CET 2004
On Tue, 02 Mar 2004 06:20:48 -0500
Tom Allison wrote:
> I'm assuming that my gaping hole in the middle is from the min_dev
> setting ( 0.1 ).
Yes. With my site's min_dev of 0.435 the hole is considerably bigger
:-)
> I'm surprised as how even the distribution is on these words.
You've got 98 low scoring words and 49 high scorers. I'd be disinclined
to call a 2::1 ration "even".
> X-Bogosity: Unsure, tests=bogofilter, spamicity=0.292113,
> version=0.17.2
> int cnt prob spamicity histogram
> 0.00 20 0.019998 0.005010 ####################
> 0.10 10 0.165554 0.024038 ##########
> 0.20 32 0.248914 0.094828 ################################
> 0.30 36 0.352749 0.177986 ####################################
> 0.40 0 0.000000 0.177986
> 0.50 0 0.000000 0.177986
> 0.60 20 0.645291 0.260171 ####################
> 0.70 5 0.759773 0.282646 #####
> 0.80 9 0.848420 0.328154 #########
> 0.90 15 0.978912 0.445043 ###############
>
>
> X-Bogosity: Unsure, tests=bogofilter, spamicity=0.292113,
> version=0.17.2
> n pgood pbad fw
> U
> "rcvd:userid" 113 0.054405 0.000000 0.000037 +
>
>
> Just for my education:
> n: number of times this word has been seen
> pgood/pbad - probably for token.
> fw - no clue...
> U - no clue...
Read the FAQ. The info is there.
More information about the Bogofilter
mailing list