Spam Assassin: Difference between revisions
No edit summary |
|||
| Line 72: | Line 72: | ||
Bayesian filtering lets spam assassin "learn" to recognize spam. You must manually sort e-mail into separate spam, and ham (any mail that's not spam) folders. Build up several thousand of each, then run the below commands. (You can run them monthly, or whatever, but it will take several thousand of them before Bayesian filtering becomes effective.) | Bayesian filtering lets spam assassin "learn" to recognize spam. You must manually sort e-mail into separate spam, and ham (any mail that's not spam) folders. Build up several thousand of each, then run the below commands. (You can run them monthly, or whatever, but it will take several thousand of them before Bayesian filtering becomes effective.) | ||
<code>sa-learn --spam /path/to/spam/folder</code> | <code>sa-learn --spam /path/to/spam/folder</code><br /> | ||
<code>sa-learn --ham /path/to/ham/folder</code> | <code>sa-learn --ham /path/to/ham/folder</code> | ||
== References == | == References == | ||
Revision as of 20:03, 20 February 2017
cPanel Settings
Spam Score Limit
SpamAssassin screens each incoming email message and scores it based on its spam characteristics. By default, SpamAssassin considers messages with a score of 5 to be spam; you can adjust the spam score limit to more or less aggressive. Select the desired spam score limit from the "Score" drop-down menu in the Filters section. The lower the score, the more restrictive the filter will be:
- 0 means everything incoming will be marked as spam.
- 5 is the default setting (and works well for typical users).
- 10 means that any message with a score of 10 or less will not be marked as spam.
Spam Box
When Spam Box is enabled, a spam folder is created. Spam is then delivered directly to this folder, allowing you to review all mail marked as spam before it is deleted.
To enable Spam Box, click "Enable Spam Box" in the Spam Box section. (This is the recommended setting, instead of auto-deleting spam.)
email filters
Filter Rules and Criteria:
Spam Status Whether Apache SpamAssassin™ marked the message as spam. The Spam Status line begins with Yes or No.
Spam Bar The content of the Spam Bar header that Apache SpamAssassin generated for this message. The more plus signs (+) that Apache SpamAssassin assigns to a message, the greater the likelihood that the system marks the message as spam.
Spam Score The total number of plus signs (+) in the Spam Bar value, expressed as an integer.
For more information about the Spam Score option, read the Spam Score section below.
Operators
begins with The message begins with the defined string.
contains The message a string that you define.
does not begin The message does not begin with the defined string.
does not contain The message does not contain the defined string.
does not end with The message does not end with the defined string.
does not match The message does not exactly match the defined string.
ends with The message ends with the defined string.
equals The message exactly matches a defined string.
matches regex The message matches a regular expression that you define. (Note: The filter text box accepts regular expressions when you select this option, rather than commonly-used wildcard characters (for example, * or ?).)
Spam Score
The following options are only applicable when you select the "Spam Score" option:
is above (#s only) The message's Spam Score is greater than the number that you define.
is not above (#s only) The message's Spam Score is equal to or less than the number that you define.
is below (#s only) The message's Spam Score is less than the number that you define.
is not below (#s only) The message's Spam Score is greater than or equal to the number that you define.
Bayesian Filtering
Bayesian filtering lets spam assassin "learn" to recognize spam. You must manually sort e-mail into separate spam, and ham (any mail that's not spam) folders. Build up several thousand of each, then run the below commands. (You can run them monthly, or whatever, but it will take several thousand of them before Bayesian filtering becomes effective.)
sa-learn --spam /path/to/spam/folder
sa-learn --ham /path/to/ham/folder