Methods

Whitelists

Only let through certain mail (e.g. from known senders)

Heuristics

Use rules of thumb to guess which mails are spam

Blacklists

Block mail from specific senders or with specified contents

Challenge-Response

Ask senders to perform a task (e.g. read wonky letters) before letting mail through. Success puts you on the whitelist.

Reviews

• Free to try Windows options - http://www.forbes.com/free_forbes/2003/0623/136.html
• Mc Afee Spam Killer - http://www.smallbiztechnology.com/smallbizarticles/spamkiller.shtml
• Spam Assassin, Surf Control, Bayesian Filters and Neural Networks - http://www.nwfusion.com/buzz/2003/0929spam.html
• 16 Products - http://www.nwfusion.com/reviews/2003/0915spam.html

Summaries

My Favourite

PopFile

Learns to classify your email into different categories, so as well as detecting Spam, it can help you sort email into work and personal etc. It uses Bayesian statistics to learn from examples you give it. Initially it regards all email as "Unclassified", you correct it by saying "this one is spam") and it very quickly learn to label your mail.

Pop File works either on a windows PC or under Linux or other operating systems. It can be installed on your client machine (i.e. desktop) or on the email server.

http://popfile.sourceforge.net/

Windows Client Software

• Cloudmark Spam Net - Community of users identify spam
• Ella - Learns from examples of spam and non-spam - http://www.openfieldsoftware.com
• GWAVA - http://www.beginfinite.com/
• I Hate Spam - http://www.sunbeltsoftware.com
• Mc Afee Spam Killer - Rules based. New rules shipped regularly.
• Mail Frontier Matador - Multiple techniques
• PopFile - Learns to recognise spam - http://popfile.sourceforge.net/
• Qurb - Uses "Whitelists"
• Spam Butcher - Pre-trained, claims to only required occasional updates to stay effective - www.spambutcher.com

Server Based

• PopFile - Learns to recognise spam - http://popfile.sourceforge.net/
• Spaminator
• Spam Assassin - http://www.spamassasin.org/
• Surf Control

Unknown

• Declude Junk Mail