Spam Filter ISP Support Forum

  New Posts New Posts RSS Feed - bayesian filter
  FAQ FAQ  Forum Search   Register Register  Login Login

bayesian filter

 Post Reply Post Reply
Author
ImInAfrica View Drop Down
Groupie
Groupie
Avatar

Joined: 27 June 2006
Location: FL, USA
Status: Offline
Points: 60
Post Options Post Options   Thanks (0) Thanks(0)   Quote ImInAfrica Quote  Post ReplyReply Direct Link To This Post Topic: bayesian filter
    Posted: 25 August 2011 at 5:06pm
Hello,
somehow my Bayesian filter has stopped filtering/recognizing spam.
what is the correct procedure to reset it, and how can we 'teach' it spam?

thank you in advance.

Back to Top
LogSat View Drop Down
Admin Group
Admin Group
Avatar

Joined: 25 January 2005
Location: United States
Status: Offline
Points: 4068
Post Options Post Options   Thanks (0) Thanks(0)   Quote LogSat Quote  Post ReplyReply Direct Link To This Post Posted: 25 August 2011 at 10:03pm
Hi ImInAfrica,

To reset the Bayesian database you can:
  • stop SpamFilter
  • delete (or rename) the SpamFilter/corpus directory
  • restart SpamFilter
The statistical engine kicks in after 5,000 non-spam and 5,000 spam emails have been received (these "learning" values are customizable by editing the SpamFilter.ini file).
Roberto Franceschetti

LogSat Software

Spam Filter ISP
Back to Top
ImInAfrica View Drop Down
Groupie
Groupie
Avatar

Joined: 27 June 2006
Location: FL, USA
Status: Offline
Points: 60
Post Options Post Options   Thanks (0) Thanks(0)   Quote ImInAfrica Quote  Post ReplyReply Direct Link To This Post Posted: 26 August 2011 at 12:14pm
thanks Roberto. how can we go about teaching it a little quicker then just waiting for emails?
Back to Top
LogSat View Drop Down
Admin Group
Admin Group
Avatar

Joined: 25 January 2005
Location: United States
Status: Offline
Points: 4068
Post Options Post Options   Thanks (0) Thanks(0)   Quote LogSat Quote  Post ReplyReply Direct Link To This Post Posted: 26 August 2011 at 4:16pm
In the SpamFilter.ini file, the value:
MinEmailsForBayesKickIn=5000
is used to determine the minimum number of "good" and the minimum number of "spam" emails that must be received before the bayesian filter starts to stop spam. You could lower those values, but doing so may decrease the accuracy of the filter. 
As the email "types" that reach each company is different on average than another (a hospital will receive different emails than a school does) you *must* receive actual emails to train the bayesian filter there are no shortcuts, sorry!
Roberto Franceschetti

LogSat Software

Spam Filter ISP
Back to Top
yapadu View Drop Down
Senior Member
Senior Member


Joined: 12 May 2005
Status: Offline
Points: 272
Post Options Post Options   Thanks (0) Thanks(0)   Quote yapadu Quote  Post ReplyReply Direct Link To This Post Posted: 28 August 2011 at 1:54am
Originally posted by LogSat LogSat wrote:

As the email "types" that reach each company is different on average than another (a hospital will receive different emails than a school does) you *must* receive actual emails to train the bayesian filter there are no shortcuts, sorry!


We do not use or offer bayesian filter to our customers, in part because of what you have listed here.  There is currently no way to make it domain specific so the usefulness is reduced.
--------------------------------------------------------------
I am a user of SF, not an employee. Use any advice offered at your own risk.
Back to Top
 Post Reply Post Reply
  Share Topic   

Forum Jump Forum Permissions View Drop Down



This page was generated in 0.113 seconds.