Spam has very rapidly grown to be more than just an unpleasant presence. It currently
costs big companies billions of dollars annually because of employees loss of productivity.
An evolutionary model for spam detection is proposed. It evolves a set of rules that induced the
membership of e-mails from a training set either to the ’spam’ or the ’ham’ category. Rules are
self adapting for their corresponding category through the means of a special evolutionary metaheuristics
called Genetic Chromodynamics. Detected rules are used to classify previously unseen
e-mails. Proposed technique may be used for more general purposes as system identification,
text categorization etc.