Can I tell Spamnix to block messages based on specific words in the message body?
Blocking messages based on the presence of specific individual words is not a useful spam-blocking strategy. Instead, you should train Spamnix on every message it mis-classifies. Doing so effectively teaches Spamnix how to recognize nearly every “indicates spam” or “indicates non-spam” word in the messages you receive, all at once, instead of just the few words you would program manually. Why do we say that blocking messages based on individual words is not useful? Yes, perhaps every message you receive with the word “viagra” is spam. However, spammers frequently misspell words in both the subject and the body of messages. Also, words can be “spelled” using foreign-language characters, special “digraph” HTML sequences, and via a number of other tricky obfuscating methods. It would be impossible for you to program Spamnix to recognize them all. So, you might create a rule to block “viagra” and then, when a message with the word “viagrà ” slipped through, you would feel that Spamnix faile
