Study and Comparative Analysis of Various Image Spamming Techniques

Abstract

Image spamming is a recently emerging technique that has gained attention because of the filtering problems faced in multimedia data. To compete with the developments in the area of spam filters, spammers started using image spam which resulted in ineffective text analysis of email body. The spam message is embedded into the attached image that is mostly modified randomly to elude signature-based mechanism of detection, and it is made obscure from OCR (Optical Character Recognition) text recognition tools. In this paper, an attempt has been made to study image spam and various image spam filtration techniques. Further, we have discussed different image spam detection techniques applied by researchers along with their pros and cons. Survey suggests that image spam has increased the complexity of spamming. Deployment of spam filtering techniques is mandatory for productivity and integrity of business system. Several approaches followed by research give positive aspect in spamming but there are certain issues with these techniques. Hence, there is a need to have hybrid approach that is cost-effective and feasible.