Evolutionary support vector machines for spam filtering

As the number and threat of unsolicited commercial e-mails seems to be increasing day by day, the need for powerful techniques that are able to combat the intrusion of these e-mails is immense. In this respect, present paper proposes a recently developed technique born through hybridization between the successful learning paradigm of support vector machines and the strong optimization potential of evolutionary algorithms. Suggested learning technique is applied to a large set of data containing information about 4601 e-mails; for each e-mail there are records about frequencies of words and other relevant data adding up to a total of 57 attributes. Obtained results encourage further investigation.