One of the best tools, used for fighting against spam emails is Apache SpamAssassin. SpamAssassin utilizes a plethora of algorithms, rulesets, and the configured Spam Threshold Score to analyze email to determine the likelihood that it is spam. In this guide, you can learn how to determine an effective Spam Threshold Score.
How SpamAssassin Scores Emails
When an email is received, SpamAssassin scans it and runs tests against the preconfigured ruleset and even custom rules. As SpamAssassin proceeds through each rule, it keeps a running total of all the points based on the results of each test. Each rule can delegate either positive or negative points to the total running score. Using the sum of all the points, SpamAssassin classifies the email as spam only if it exceeds the configured Spam Threshold Score.
NOTE: SpamAssassin does not block emails based on the score, rather it will append a standard string of text (***SPAM***) to the headers of the email. This is how the Spam Filters configured settings dictate how the message should be handled.
Choosing a Spam Threshold Score
The Spam Threshold Score plays an important role in classifying emails as legitimate or spam messages. Unfortunately, one size does not fit all when it comes to setting the optimal value for classifying spam. The best way to find the most effective threshold score is trial and error.
First, you should enable the Spam Box and disable the auto-delete features in the Spam Filters settings. This will filter emails, that exceed the threshold score, to a dedicated folder (for you to further review) rather than automatically deleting them. This also prevents legitimate emails from being automatically deleted if the threshold score is too strict.
Once that is done, you can proceed to set the Spam Threshold Score. At first, you can try to use the default value (5). With these settings in place, you should monitor the spam box and inbox closely for a few days to a week (depending on the amount of emails you receive). If you see more spam emails being sent to your inbox, you should change your threshold score to a lower value (to be more strict). If you see more legitimate emails being sent to your spam box, you should change the threshold score to a higher value (to be more lenient).
Although taking an active approach to controlling the amount of spam you receive may seem daunting, it can be as simple as trial and error. The suggestions above will help to make it is easy to find that Goldilocks, “just right” setting though.