Adaptable Text Filters a and Unsupervised Neuralnd Classifiers for Spam Detection

摘要

Spam detection has become a necessity for successful email communications, secu- securityand convenience. This paper describes a learning process where the text of incoming emailsrity is analysed and filtered based on the salient features identified. The method described haspromising results and at the same time signifi significantly better performance than other statisticalcantly and probabilistic methods. The salient features of emails are selected automatically based onfunctions combining word frequency and other discriminating matrices, and emails are thenencoded into a representative vector model. Several classifiers are then used for identifyingspam, and self-organising maps seem to give significantly better results.