• Top
• Download
• Documentation
• Mailing Lists
• Bugs, Patches, and Requests
• CVS
• SourceForge Project Page

Steps to block spam with Pyzor

Introduction

This document describes general overall steps you should take to use Pyzor effectively.

Steps

1. Blacklist

   The first step you should take on your incoming mail is to process it with a blacklist. A blacklist blocks all mail that can be definitively and deterministically marked as spam. A common technique is to keep a list of commercial spammers and block them at this step. For example, you may want to block all mail coming from mediatrec.com, since they are spammers; they would then be blacklisted.

2. Whitelist

   The second step you should take on your incoming mail is to process it with a whitelist. A whitelist allows in all mail that can be definitively and deterministically marked as coming from a "good source". This includes mailing coming from personal people you know and mailing lists.

3. Content-based solutions

   Content-based solutions look at the body of an email, and use methods that are not definitive and heuristics-based to mark things as spam.

   It is vital that you use content-based solutions after whitelisting, since they can mark 'good' mail as spam. Any system which uses heuristics or non-definitive methods should always uses these methods as a fallback from the more assured blacklist and whitelist. To use these systems with missing, incomplete, or incorrect whitelists is to show haphazard care for marking 'good' mail as spam.

   Examples of content-based solutions are:

   • Pyzor, which calculates digests of parts of mail, and matches it with what other people have gotten.
   • SpamAssassin, which uses heuristics such as looking for many caps along with a scoring method to mark things as spam. Note that products such as SpamAssassin often use products such as Pyzor as part of their tests.