Image Spam Detection Through Server-Client Filtering by Tracing the Source IP of the Spammer

Image spam is a type of e-mail spam that embeds spam text content into graphical images to bypass traditional text based e-mail spam filters. To effectively detect image spam, it is desirable to leverage image content analysis technologies. A solution to determine spam images is to embed both server-side filtering and client-side detection. On the server-side, spectral clustering algorithm is used and in the client-side detection, active learning principle is used. By using spectral clustering algorithm on the server side, it will cluster the spam images and filter the attack activities of spammers and fast trace back the spam source. By using active learning on client-side filtering, the learner guides the users to label as few images as possible while maximizing the classification accuracy. The server-side filtering identifies large image clusters as suspicious spam sources and filtering method is performed to identify the real sources and block them from beginning.