fedora core 5 problems

[David Relson]

On Wed, 7 Jun 2006 21:37:18 -0700
OTR Comm wrote:

> hello all,
> 
> it has been a number of years since i have monitored this list, so if
> i ask some redundent questions, please forgive.
> 
> i have fedora core 5 system, kernel 2.6.15-1.2054_FC5smp, berkeleydb
> 4.3, gsl-1.8, and trying to get bogofilter 1.0.2 operating.
> 
> i can get it to compile fine, but it is not classifying messages like
> i specify.  i like to start with a clean slate without using a
> corpus, so i start with a single message.
> 
> i am on my system as root,and sitting in the directory where i have a
> single message, then i give the command:
> 
> cat * | bogofilter -n -v
> 
> and the response is:
> 
> Created directory /root/.bogofilter .
> # 49 words, 1 message
> 
> good so far.  then i give the command:
> 
> bogofilter -e -p -d /root/.bogofilter -B * | grep Bogosity | grep
> spamicity and i get:
> X-Bogosity: Unsure, tests=bogofilter, spamicity=0.520000,
> version=1.0.2
> 
> and this is not correct since i had initially told bogofilter to
> register the message as ham.
> 
> could somene please tell me what is going wrong here?
> 
> btw, i have bogofilter 0.90 that i installed three years ago running
> on other redhat boxes, and have 0.90 runnng under cygwin on some
> windows boxes (nt, xp, and 2k), but i can not get 1.0.2 running under
> fedoar core 5.

For initial training, give bogofilter 2 messages - 1 ham and 1 spam.
It'll be happier.

In your test, you have only a single ham in the wordlist, so every
token of the incoming message will either match a token in the wordlist
or be an unknown. There's no realistic way for bogofilter to produce a
sure result.  

As a test I created a directory with 2 ham (named h1 and h2) and 2 spam
(s1 and s2).  Then I trained with 1 ham and 1 spam, then scored the 4
messages.  Here are the commands and the results:

#!/bin/sh
ls -l
bogofilter -C -d . -n -B h1
bogofilter -C -d . -s -B s1
bogofilter -C -d . -v -B h1 h2 s1 s2

[relson at osage FC5]$ ./test.sh
total 36
-rw-rw-r--  1 relson relson  1556 May 19 20:21 h1
-rw-------  1 relson relson  2057 Jun  8 02:03 h2
-rw-------  1 relson relson  1310 Jun  7 23:26 s1
-rw-------  1 relson relson 17412 Jun  8 04:15 s2
-rwxrwxr-x  1 relson relson   109 Jun  8 07:20 test.sh
h1 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.0.2
h2 X-Bogosity: Spam, tests=bogofilter, spamicity=0.999288, version=1.0.2
s1 X-Bogosity: Spam, tests=bogofilter, spamicity=1.000000, version=1.0.2
s2 X-Bogosity: Unsure, tests=bogofilter, spamicity=0.826015,
version=1.0.2

HTH,

David