Someone makes an obvious discovery about Bayesian filtering the very long way round and it warrents a BBC News article. Geeze. No wonder these worms find enough stupidity to spread.
Update: Yes, I know it's the POPFile author. Doesn't mean that he's not being an idoit. He could just have opened the Bayesian db to check which words had a high positive value. Probably the BBC journalist had something to do with the crapness however.