A model for selective attention in monitoring and control reasoning tasks

The author describes a selective attention model which is based on the development of future expectations and contingencies in planning. Selective attention is the ability to intelligently direct the reasoning/planning focus in order to dynamically respond to changing events in monitoring a physical system. The selective attention mechanism enables active sensing which is employed to confirm or deny hypotheses related to future expectations. Contingency plans are modified accordingly. A program was developed based on this model. The program is part of ELICIT (expertise learner and intelligent causal inference tool), coded in Joshua (a high-level representation language developed by Symbolics, Inc.) on a Symbolics 3630 LISP machine