Research on semi-Naive Baysian classification model in anti-spam
Ping Dong · Journal of Wuhan Polytechnic University · 2011
In order to improve the precision and efficiency of spam filter,this paper gives a N-Average on One Dependence Anti-Spame Filter pattern based on x-dependence Bayesian classifier.This model uses N-one-dependent Baysian classification to estimate the average probability of the type of mail to discriminate.The experiment proves that this method can greatly improve the classification accuracy and get lower classification error rate,and the classification rate is also satisfactory.