Computer workstations: Explorations in human and machine cognition

Professional and scientific work will increasingly be conducted at computer workstations. An expert-system adaptive interface is one method for supporting learning and raising productivity of workstation activities. A prototype application for an adaptive interface is described in terms of machine intelligence, needs, design philosophy, and cognitive research. The intelligent interface prototype includes the operator’s psychometric profile, prior performance history, and a current keystroke performance stack within a conventional expert system. Productions in the form of conditionals between the operator’s keystrokes and three types of commands to the underlying software are described and illustrated. Commands entered, syntax errors, excessive repetition of keystrokes, and skill level are used to detect the user’s intentions and accordingly modify the display and command structure of the underlying software. Determining human intentions to produce machine adaptation to individual users is not a trivial task. Efforts to develop a powerful and general adaptive interface pose significant problems, and success cannot be assured without intensive cognitive research.