Soft Computing-Based Information Security

Abstract

This chapter deals with using soft computing methods in information security. It is engaged in two big areas: (1) information security and spam detection and (2) cryptography. The latter field is covered by a proposal of an artificial neural network application, which represents a way of further development in this area. Such a neural network can be practically used in the area of cryptography. It is a new approach, which presents a development of automatic neural networks design. The approach is based on evolutionary algorithms, which allow evolution of architecture and weights simultaneously. A spam filter is an automated tool to recognize spam so as to prevent its delivery. The chapter contains a survey of current and proposed spam filtering techniques with particular emphasis on how well they work. The primary focus is spam filtering in email, but the role of the spam filter is only one component of a large and complex information universe. The chapter also includes experimental demonstrations.