Anti-spam - Tutorials
These tutorials details how to set up the Mailtraq intelligent anti-spam system. To learn more about the principles behind the system read about How it works before you get started.
The anti-spam system is a learning system - it isn't pre-programmed with a set of rules (which would quickly go out-of-date), so either the administrator, or individual users, need to teach it what they think is spam and what is regular mail.
- If the individual users do this, they can train the system to their personal preferences, and it will assess their incoming mail based on those prefences, giving them the 'best' results.
- If the administrator does this, all the users' mail will be subject to the same assessment. The users won't need to be involved in the anti-spam training process, but will be reliant on the administrator to correctly spot what is regular mail during the training phase.
All the information the system collects about spam and regular mail is stored in a 'database'. A database is a 64Mb file held on the Mailtraq server, so if each user will be teaching their own databases you need to ensure there is sufficient disk space for all the databases.
Tip: The best defence against spam is to adopt a multi-layered approach to spam control. We recommend using a DBL, together with this Bayesian anti-spam system with SpamAssassin enabled.