A Discriminative Classifier Learning Approach to Image Modeling and Spam Image Identification

The authors propose a discriminative classifier learning approach to image modeling for spam image identification. This paper analyzes a large number of images extracted from the SpamArchive spam corpora and identifies four key spam image properties: color moment, color heterogeneity, conspicuousness, and self-similarity. These properties emerge from a large variety of spam images and are more robust than simply using visual content to model images. The authors apply multi-class characterization to model images sent with emails.