I've noticed a lot of spam getting through at the moment with links to docs.google.com These emails are all random text based, which makes it very hard for bayes to do anything with it. Is anybody effectively detecting these as spam? If so, which rules are generally picking them up?