Developing a Computer Model of Human Personalities Suitable for Robotics and Feedback Applications

This paper presents a quantitative, predictive model of human self-images consistent with several personality temperament indicators including Lowry Colors, Enneagram, Merrill's C.A.P.S., as well as with the clinical personality disorders. We predict how the Myers-Briggs conflict-type behaviors stem from this model, and validate the ideas with a computer simulation that closely predicts the observed distribution between the Enneagram and Myers-Briggs conflict types in the human population. The theory directly results in the ability to design feedback systems with personalities. It also clearly points toward how to develop robots with human-like behaviors and chat-bots with human-like personalities.