Title
Bayesian Spam Classification Applied to Phishing E-Mail
Author / Source
SonicWALL, Inc.
Description
Effective e-mail security solutions must protect users from all types of e-mail threats. However, each threat must be treated differently to address its unique properties. Bayesian filtering has been successfully combating spam. However, this same filter cannot be applied to phishing. Instead, a Bayesian phishing filter can be created through layering different features which are strong indicators of phishing. Test results show that Bayesian rules can be effectively applied to identifying phishing. This identification allows phishing e-mails to be uniquely categorized. Even if recipients can not recognize the e-mails as phishing, the e-mails can be securely removed from the inbox and labeled appropriately, helping to keep users safe.
File Information
15-Sep-2008 / English / US Letter Size / 0.15 MB / PDF
