Model-based troubleshooting for information survivability
Howard E. Shrobe · 2002
To be survivable a system must be adaptable, able to respond to attacks and to achieve its goals. This paper considers one approach to adaptability, self-diagnosis and repair. We explore how ideas from the artificial intelligence community in model-based troubleshooting can be applied to this task and present a new method for self diagnosis which combines logical and statistical modeling and reasoning.