Affect-Aware Conversational Agent for Intelligent Tutoring of Students in Nursing Subjects

In many social professions employees require skills in affect- and situation-aware social interaction. One option for teaching and training such social interaction skills by computer-based training methodology is the use of dialogue simulations. Here, a student interacts with a simulated dialogue partner and the dialogue flow explores specific interaction situations and affectual settings. Conversational agents provide a basic technology for creating such dialogue simulations. However, they usually lack a means for managing affect-related dialogue state. In this paper we propose an approach to integrate affective reasoning into a conversational agent for intelligent tutoring applications in order to improve the agent’s ability to recognise dialogue intents, generate emotionally aligned responses, and provide a metric for evaluating student performance.