GETTING.STARTED (rev 2)

[David Relson]

On Wed, 27 Oct 2004 17:21:05 +0200 (CEST)
Boris 'pi' Piwinger wrote:

> David Relson said:
> 
> >> > Eh??  Bogotune uses the wordist, and the ham and spam corpora you
> >> > specify, and then does a rather exhaustive scan of possible
> >scoring> > parameters to find what gives the best results.  As you
> >know,> > bogotune has minimum requirements for number of messages
> >registered> > in wordlist.db and minimum numbers of messages for the
> >ham and spam> > corpora used in the tuning process.
> >>
> >> Right, so it is not usable for pure train-on-error approaches.
> >
> > Usually when I run bogotune, I start with an empty wordlist and
> > 10K-15K ham and 10K-15K spam.
> 
> This is one approach. Others are listed in the FAQ. Also some people
> don't have huge mail archives.

True.  My archive is larger than most.  In any case, bogotune needs to
have a large set of messages for what it does and expects a large number
of messages to be represented in the word list.  

I'm thinking of a toe-tune (train-on-error + bogotune) experiment which
will probably use a modified bogotune that doesn't care about the
message count in the wordlist.  It'll be interesting to see if bogotune
runs happily or something goes wrong.