An efficient spam filtering method by analyzing e-mail’s header session only
許志堅, Jyh‐Jian Sheu, Ko‐Tsung Chu · International journal of innovative computing, information & control · 2009
This research manages in-depth analysis on Chinese spam and expects to find more efficient spam judgment rules in the condition of reducing system burden. Different from many spam filtering methods at present which have to investigate the complete content of e-mails, this research doesn’t check the content of e-mails to avoid the complexity and executive efficiency. We focus on header’s basic attributes such as e-mails titles, senders’ names, senders’ e-mail address, sending date and apply decision tree data mining technique to analyze the association rules of Chinese spam and propose a systematic method with reversing mechanism to accurately identify spam and legitimate mails. According to the experiment, the accuracy of our spam filtering method proposed in this paper was 96.17% and the precision of our method was up to 98% which was not lower than that of other present filtering methods of checking e-mail content. Thus, the method of this research could efficiently identify the spam e-mails by only checking the header sections. This advantage of this method could reduce the cost for calculation.