Abstract
Partially observable Markov decision process (POMDP) model has been demonstrated many times to be suited for developing robust spoken dialogue systems unreliable speech recognition. In this paper, we propose a new factored POMDP model to describe a new application on building affective tutoring system (ATS). Different from previous models, the user's state space is divided into three components: goals, dialogue states, and emotions. Moreover, the system's action space is factored into two parts: goal response and emotion response, in order to respond to the user's goal and emotion, respectively. We further describe how to apply the proposed model to build an ATS in detail. Five experiments are designed to reveal the influence of some key parameters on the system performance, and the simulation results demonstrate the validity and feasibility of the proposed model. © 2018 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
