Using optimize na"ive Bayes classifier in problem of SMS classification

In article the process of optimizing a naive Bayesian classifier is viewed. The mechanism of calculating the probability and the process of constructing a training table are analyzed. It’s needed for a practical solution to problem of SMS messages classification by Naive Bayes classifier. There is experimental set of SMS messages. It used by experts for SMS messages classification process. It has two sets - train set and test set. The quality of SMS messages classification is higher for optimized method and lower of non-optimized Naive Bayesian classifier.