For the WordPress users out there there is a write up on doing spam post blocking.
I'm on the fence about Bayesian filters and content spam. The online casino folks who are spamming away the last few days use text that is not necessarily what I want to teach my filter to sort out. It would be a bit better if it matched content with the post to see if text from the Gettysburg address is appropriate to the context. The filtering of URL's and words in content spam prevention seems to be more useful so far.