Frances L. Johnson
Johnson is revising the traditional belief revision integrity constraints and postulates so that they can be applicable to the deductively open belief space of any implemented system. Current theories consider belief revision in the context of a deductively closed belief space, but real, implemented systems cannot afford to be deductively closed. After developing belief revision theories for a deductively open belief space, Johnson will implement these as enhancements to an existing KRR system and use them as guidelines to evaluate the quality of belief revision in various systems.