Abstract

This paper examines how the momentum of emotion can be modeled and reflected within expressive Internet communication systems. We describe a prototype real-time emotional communicator. A core component is the text-to-emotion extraction engine. The engine analyses the text that is input by users, extracts the emotion that is contained and outputs the parameters that are necessary to invoke an appropriate image expressing the emotion. The parameters include both the detected emotion and the intensity. The concept of ‘Emotional Momentum’ is incorporated in our engine to resolve ambiguity or conflicting emotional content that occur within individual sentences.