Naive Bayesian classifier adaptation for e-mail classification mechanism

Actually there are many difficulties for solutions of email classification problems. One is the problem of content analysis for two classification groups containing reliable and unreliable data. There are known a number of adaptive and non-adaptive algorithms that should help to solve described problem. Nowadays naive Bayesian classifier algorithm is one of the most popular tool in the field of data classification problem solution. This work is concerned on how to adapt naive Bayesian classifier mechanism for e-mail classification, where e-mails are classified as reliable and unreliable information blocks. We determine naive Bayesian classifier learning process as calculation the probability of one or another word meeting into e-mails.