Research of a Spam Filtering Algorithm Based on Naïve Bayes and AIS

The Naïve Bayesian classifier has been suggested as an effective method to construct anti-spam filters for its strong categorization and high precision. Artificial immune system has become a new embranchment in computing intelligence for its good self-learning, self-adaptability and robustness. This paper proposes a new spam filtering means based on Naïve Bayes and AIS, and analyses the key problems of the algorithm. The accuracy rate is compared with a naïve Bayesian classifier-Bogofilter and it is shown that the proposed algorithm performs as well as Naïve Bayes and has a great potential for augmentation.