How are the worst offenders identified
A series of litmus tests has been defined, each test looks for different clues that may (when seen together) make it obvious that a message really is spam. for example, the IP test looks for multiple messages that come from the same IP address.
What litmus tests are there?
- IP Address - multiple messages from one source.
- MD5 - multiple messages where the body text has the same MD5 hash
- Link Count - messages that contain more than n Links.
- Domain - Messages advertising the same URL
- Email - multiple messages from one email address.
- Name length - messages where the author name has more than 5 words
Can more litmus tests be devised?
Yes, the test system is plugabble.