Main Article Content
The rapid spread and the easy availability of a free e-mail service have made it the medium of choice for the sending
of unsolicited advertising and bulk e-mail in general. These messages are called spam e-mail and are an increasing problem to
both Internet users and Internet Service Providers (ISPs). In this paper, a multi-agent based architecture is adopted to allow for
efficient on-line application of the developed filter while maintain the scalable design that lend itself to openness, a multi-agent
system (MAS) designed to implement a spam filter system based on mobile agent technology with SVM-filter that works
background, at client side, to improve classification accuracy of Gmail and to allow for distant classification.