Development of a Conversational Agent for Tutoring Nursing Students to Interact with Patients

Tomasz Sosnowski, Moh'd Abuazizeh, Thomas Kirste, Kristina Yordanova

Abstract: Conversational Intelligent Tutoring Systems can provide a more interactive and engaging learning experience compared to classical teaching methods. As such, conversational tutoring agents have their place as an additional learning resource for students, especially when a need arises for a student to work through a learning scenario which would be difficult or dangerous to reproduce in real life. Such is the case with the training of caretakers for nursing homes and hospitals, needing to train nurse-patient interaction. In this work we present our approach to the development of a conversational agent intended to simulate a patient. The conversational agent is flexible enough to handle large number of possible conversations with the help of automatic story generation methods. This approach enables us to deploy new conversational agents with no need for manual writing of every possible storyline, while remaining compatible with available commercial off-the-shelf chatbot frameworks. Our work was concluded with a user study validating positive usefulness of this type of a system.