基于加权朴素贝叶斯的邮件过滤方法

In allusion to MI feature extration algorithm and naive bayes algorithm, this paper introduces the concept of the Feature Term Discrimination(FTD),analyse the discrepancy of the separating capacity of various feature terms in the categorizing process, and put forward a kind of feature extraction algorithm which give consideration to both FTD and MI. By further add FTD to the design of classification algorithm,a weighted na?ve bayes algorithm is presented to solve the problem of content-base filtering efficiently.The experimental results show that the improved algorithm has increased significantly in terms of the recall rate, precision rate and accuracy rate , and the performance of classification is more stable.