摘要

Based on the Pawlak rough set theory, this paper investigates separations in covering approximation spaces, give some characterizations of these separations and some relations among these separations. As an application of these results, investigations on network security are converted into investigations on separations in covering approximation spaces by taking covering approximation spaces as mathematical models of networks. Results of this paper give further applications of the Pawlak rough set theory in pattern recognition and artificial intelligence, which makes it possible to research network security by logical methods and mathematical methods in computer science. This contributes to giving risk assessments of securities and to raise grades of securities for networks.