Image Spam Filtering by Content Obscuring Detection

TitleImage Spam Filtering by Content Obscuring Detection
Publication TypeConference Paper
Year of Publication2007
AuthorsBiggio, B, Fumera, G, Pillai, I, Roli, F
Conference NameFourth Conference on Email and Anti-Spam (CEAS 2007)
Date Published02/08/2007
Conference LocationMicrosoft Research Silicon Valley, Mountain View, California
Keywordsdoc00, doc02, spam filtering

We address the problem of filtering image spam,a rapidly spreading kind of spam in which the text messageis embedded into attached images to defeat spam filtering techniquesbased on the analysis of e-mail's body text.We propose an approach based on low-level image processingtechniques to detect one of the main charactersticsof most image spam, namely the use of content obscuring techniquesto defeat OCR tools.A preliminary experimental evaluation of our approach is reportedon a personal data set of spam images publicly available.

Citation Key 139
biggio_CEAS2007.pdf486.14 KB