Publisher's Synopsis
"Critiquing Human Error" provides design guidelines for producing successful expert critiquing systems and presents COPE, a critic research testbed. It also unifies the two traditions in the study of human expertise which have been directed towards reducing expert error and judgement bias: the expert systems model and the judgement and decision making model.;The author, with his colleagues and students, has constructed five reusable critic designs to date, covering generic cognitive tasks such as forecasting, planning design, authoring, knowledge base generation and probabilistic reasoning. Several of these designs co-exist in an expert critic application called TIME which as a recipient of the 1991 Innovative Application of Artificial Intelligence Award from the American Association of Artificial Intelligence (AAAI).;In a collaborative system, the human user and the computer interact as equal partners do their mutual task. To date, researchers are basing such joint cognitive systems on models from other modes of human-computer interaction, as well as attempting to adapt other models. Failures arise because these metaphors and models lack a proper theory, a set of principles, and an experimental basis for guiding the design of collaborative systems.
