An Effective Algorithm for Improving the Performance of Naive Bayes for Text Classification

Naive Bayes algorithm is uncomplicated and effective in text classification and experiments. However, its performance is often imperfect because it does not model text well, and by inappropriate feature selection and some disadvantages of the Naive Bayes itself. This paper makes some modifications for Naive Bayes to improve the performance of Naive Bayes and the effect, condition as well, on categorization. Finally, the paper adopts this algorithm in Spam Filter categorization, a quite typical text classification. Some experiments were done with this method; results were compared with its previous method.