An anti-spam architecture combining visual and semantic features

It is well known that Unsolicited Commercial Emails (UCE), commonly known as spam, are a serious problem for email accounts of single users, small companies and large institutions. The aim of our research is to define a methodology and to design an architecture in order to overcome some problems that are still boarded on the state-of-art spam-filters. The approach takes into account the semantic richness of natural languages and the spam evolution such as the use of image spam. We finally propose an experimental planning and a comparison with respect to existing tools.