Scalable Service Based Antispam Filters

Abstract

Spam messages are becoming a serious problem for most system administrators and computer users. Along with the increase in spam volume there is also a need for efficient detection methods which are able to cope with the analysis of thousands of emails per day. In this article we present a novel solution for spam detection which is based on a service oriented architecture, able to scan and classify a large number of emails. We prove that the service based architecture is scalable, capable of using a distributed system to split the computation along multiple nodes. This approach is not only very efficient and capable of handling large volumes of emails, but also very configurable, allowing the rapid change in scanning policy.