Evaluating the Error Risk of Email Filters Based on ROC Curve Analysis

Authors:  Wenbin Li, Ning Zhong, and Chunnian Liu

Filtering e-mail is a cost-sensitive task, because missing a legitimate message is more harmful than the opposite error. Therefore, how to evaluate the error risk of a filter which is trained from a given labelled dataset is significative for the e-mail filtering task. This paper surveys the researches on Receiver Operation Characteristic (ROC) curve analysis, and discusses how to use the techniques of ROC curve analysis to evaluate the risk of email filters. This work is useful for designing a bread-and-butter filter.