Data Processing and System Safety

摘要

The vulnerability of different data processing structures to accident or sabatoge as well as computer information security are considered. Characteristics of data processing architectures (centralization; decentralization; site location) and security implications are discussed. The multiplicity of structures possible in automated systems emphasizes the need for security which should be defined by precise objectives and, if possible, quantified. A system safety program that allows for risk analysis is presented. The program, PARI, is composed of 12 modules. Data processing risk is determined by an overall fault tree method. The importance of a systematic approach is stressed.