An evolutionary learning spam filter system

Accurate spam filters are of high necessity in present days as the high amount of commercial mail entering accounts has become a real threat to everyone, from causing personal computers to crash to costing big companies billions of dollars annually because of employees loss of productivity. Moreover, lately, spam also carries viruses along. Current paper presents an evolutionary model of a spam filter which can also be generalized to the more complex issue of text categorization. The model learns the rules that lie behind the classification of training e-mails into spam and non-spam and then uses them to label unseen, incoming mail. The evolutionary learning classifier system uses genetic chromodynamics to evolve the rule set. A comparison of its correctness of prediction depending on whether chromosomes representing rules are binary or real encoded is conducted. Experimental results showed that the binary encoding appeared to be less effective than the real one.