Size of Bayesian Dictionary

Discussions on webmail and the Professional version.
Post Reply
criz
Posts: 35
Joined: Thu Apr 08, 2004 2:30 pm
Location: NY

Size of Bayesian Dictionary

Post by criz » Mon Mar 24, 2008 3:44 pm

I'm using Pro v3.14 after just upgrading from 1.85. It's my first day of setting up the Bayesian filter and I have a few questions:

1) - I've got the ham filling up from authenticated users - it's around 2000 right now, and I'm assuming the Spam will start filling up (from the start up figure of approx 6392?) once the ham reaches the same amount?

(I have a filter setup to prefix subject lines of mails with spam percentage of equal/above 90%).

2) Is there a maximum dictionary size? Will the system eventually stop it growing if I just leave it on autotraining, or is there a point when autotraining should be turned off?

Thanks!

dreniarb
Posts: 316
Joined: Mon Jan 19, 2004 5:00 pm
Location: Marion, IN

Post by dreniarb » Thu May 22, 2008 11:38 am

My current ham count is at 119k, and my spam count is at 97k.

I was wondering if others had any recommendations on what the levels should be? Should we take any action to reset the dictionaries every once in a while since the nature of spam can change?

I'm curious to know how big other users dictionaries are?

Post Reply