{"body":"Bayesian filtering, summarized in the simplest sense, means that e-mails are analyzed for special strings that occur in spams and hams (legitimate e-mails). SpamAssassin is able to analyze these occurrences and help decide whether an e-mail is spam or ham through scoring based upon the Bayesian probability (percentage). A Bayesian percentage between (40,100]% will add points to an e-mails spam score, below 40% will reduce the score. When you are first starting off and training it, most e-mails will get a 50% rating (+ 0.001 points). As time progresses, SpamAssassin will learn to adapt and begin marking spams more confidently with higher ratings. A properly trained system will usually mark spam with between (80,100]%","title":"Bayesian Filtering Information"}