An Enneagram based Model for Personality Based Adaptive Systems

A model to seamlessly adapt to personalities of agents (human or artificial) in a multimodal environment is presented. The model is based on a generic personality typing technique called the Enneagram, to discover the traits stemming from egoistic stimulus. Specific traits are quietly discovered based on a hidden dialogue management system that employ a mixed initiative frame representations to guide the system towards inferring the personality type by identifying specific patterns of behavior.