Improved Nave Bayes combining feature with noncharacteristic information and its application
Xiaoli Chen · Jisuanji yingyong yanjiu · 2011
Nave Bayes algorithm was widely used in the content-based filtering,but traditional Nave Bayes faced many problems,such as the uncertainty of classifying e-mails by analyzing e-mail content,the incompleteness of e-mail representation.In order to overcome these shortcomings,this paper analyzed different attributes between ham e-mail header and spam e-mail header,extracted noncharacteristic information,and improved Nave Bayes algorithm which combined feature information with noncharacteristic information.Experimental results show that the improved Nave Bayes classification approach increases the recall and the precision of spam,covers e-mail information,and makes up for the shortage of content-based filtering,compared with that of only using feature information.