Emotional interaction model for a service robot

This paper presents the emotional interaction model that receives classified inputs from user's response and decides what emotional response the robot should generate. Cognitive emotion modeling requires profound understanding about human's cognitive processes and ideas on how to implement each constitutional components of the model into the robot. The proposed model is composed of two layers: reactive and deliberative layers. Reactive layer is in charge of immediate emotional response of the robot. It works with pre-defined rules which relate input to its corresponding emotional expressions. This layer enables immediate display of the robot's emotional state to the user so that lifelike characteristics of the robot can be achieved. The deliberative layer is in charge of the appraisal-based emotion expression and carries out the function of "action-coloring," which adds flavor to the actions being taken. Emotional interaction scenarios which are considered to be possible with the proposed emotion model are also introduced.