Combining Scheduling Heuristics to Improve e-mail Filtering Throughput

Abstract

In order to alleviate the massive increment of spam deliveries all over the world, spam-filtering service providers demand the development of new filtering schemes able to efficiently classify illegitimate content using less computational resources. As a consequence, several improvements have been introduced in rule-based filter platforms during the last years. In this context, different research works have shown the relevance of scheduling the evaluation of rules to improve filtering throughput, contributing some interesting heuristics to address this task. In this work, we introduce a novel scheduling approach that takes advantage of the combination of individual heuristics for globally improving filtering throughput.