Text classification with unbalanced classes using Bayes formula

A common method for text classification is using multinomial naive Bayes (MNB) that is a version of naive Bayes with relatively simple calculations for redicting the related classes, using a data set with outcome of high performance. In this paper, we resolve one of the problems of text classification for imbalanced data sets and also we propose a correction to previous works for adjusting their characteristics. This correction can be used as other data normalization step. Modified version of classification gives much more accurate results. The results from our experiments show that, the precision of this method is more than 0.94, which is a noticeable result in the Persian text document classification