Abstract

Network-Based Human Behavior Modeling for Adaptive Training

Chu, Y. and Madni, A.M.

Abstract

This paper presents a Modified Petri-net (MPN) characterization of human behavior for use in adaptive training. The model is shown to have a two-fold application: (1) to model the behavior of an expert or exports; (2) to model and "track" the behavior of the trainee relative to the expert model with a view to illuminating performance skill deficiencies and providing instructional feedback during, and at the conclusion of, task performance. It is suggested that trainee models can be constructed as specific "overlays" on or transformations to the expert model. The possible, transformations of the expert model within the NPN framework include stimulus or cue recognition errors, time constraints violations, precedence errors, procedural errors, and concurrency errors.

From:

Chu, Y. and Madni, A.M., Proceedings of 1985 IEEE International Conference on Systems, Man, and Cybernetics, Tucson, Arizona, November 1985, pp. 511-512.