Product manual
GFI MailEssentials 7 Anti-Spam | 128
NOTE
To find the conditions to export, use the controls under the list of conditions to move
between the pages listing the conditions.
2. In the FileDownload screen, click Save and select a folder where to save the export file.
To import conditions:
1. From the Conditions list area within the General or Subject tab, key in the folder and filename
of the file to import.
2. Click Import.
7.1.12 Bayesian Analysis
An anti-spam filter that can be trained to accurately determine if an email is spam based on past
experience.
This manual also contains information how the Bayesian filter works and how it can be trained. For
more information, refer to Appendix 1 - Bayesian Filtering (page 265).
The Bayesian Analysis filter is NOTenabled by default.
IMPORTANT
Enable learning from outbound emails and allow at least a week for before enabling
filter. This is required because the Bayesian filter acquires its highest detection rate
when it adapts to your email patterns.
Configuring the Bayesian filter
Configuring the Bayesian filter requires 2 stages:
Stage 1: Training the Bayesian filter
Stage 2: Enabling the Bayesian filter
Stage 1: Training the Bayesian filter
The Bayesian filter can be trained in two ways:
Method 1: Automatically, through outbound emails.
GFI MailEssentials processes legitimate email (ham) by scanning outbound emails. The Bayesian filter
can be enabled after it has collected at least 500 outbound emails (If you send out mainly English
email) or 1000 outbound mails (If you send out non-English email).
To do this:
1. Go to Anti-Spam > Anti-Spam Filters >Bayesian Analysis.
2. Select Automatically learn from outbound e-mails.
3. Click Apply.
Method 2: Manually, through existing email.
Copying between 500-1000 mails from your sent items to the This is legitimate email sub folder in the