This thesis designed and realized a kind of spam email filtering system on the basis of the improved Bayesian algorithm. On the foundation of the Naive Bayesian classification, the further improvement was done to the Bayesian algorithm, words of the actual meaning but not single word were participated in classifying as the characteristics, the comprehensive consideration of characteristic relevance of the word and validity was also increased at the same time, which had not only improved the accuracy of classification but also improved the efficiency of classification. The experimental results also showed a better performance in the spam email filtering with this method.
展开▼
