How emotions affect relations between interlocutors in bad news conversations

Using language involves more than exchanging information through a combination of words. Emotional elements in language can strongly influence the meaning of conversational actions which goes beyond the simple exchange of information. To evaluate current emotion models, we study how they deal with situations that arise in conversations (e.g. the intentions of a speaker). We will give insights on how computational models of dialogue can take features of emotions models into account. We present a prototype of dialogue system that allows one to practice holding bad news conversations. We discuss the ideas behind the system, the architecture, parts of the implementation and a study that has been conducted to gain more insight into the features of bad news conversations and the requirements they place on emotion models.