Towards Self-Learning and Fully Transparent UCE Prevention

A self-learning system for preventing unsolicited commercial or bulk e-mail (UCE or UBE) is presented. It acts at the source of each e-mail message and controls the traffic going out of a network, thus avoiding common drawbacks of standard spam filtering techniques. The system is based on a token bucket mechanism proposed earlier. In this paper, it is shown how to develop this approach into a fully transparent and effective UCE prevention system by introducing adaptivity and learning capabilities. The first central component introduced in this paper is a framework for quantitatively analyzing the business model underlying UCE, allowing for insights into the effectiveness of filtering in the given situation. The second one is a strategy allowing the system to adaptively and intelligently (re-)configure itself in order to achieve almost arbitrarily high levels of transparency, i. e., to harm the business model underlying UCE without affecting regular e-mail users. Finally, a third component introduces adaptability for ensuring that only spam is blocked and that no regular mail traffic is affected.