Download Antispam Marisuite 0.0.8.24

Core of the spam filter is a usual Bayesian filter. But it uses original methods for statistical analysis messages and it takes into consideration much more messages' characteristics then classical filters. Also our spam filter uses some computational approaches which allow carrying out calculations more exactly then other such programs. For filter training is used original adaptive technique, which allows automatically tune up antispam knowledge base to take into consideration new spam characteristics which are constantly changing. Our spam filter really recognizes all known methods for passing through spam filters. It filters spam, which contains advertisements in pictures, independently of distortion, pictures' format and presence or absence message text. To reduce false positives spam filter uses methods based on social networks theory. Bounces practically fully exclude probability to lose important messages. Our spam filter never deletes messages.