System-user interaction and response strategy in spoken dialogue system

Abstract

There are a number of restrictions in human-machine interactions, which continue to warrant a better con-trol of response utterances in spoken dialogue systems. Indeed, the human user often has to deal with unnatural responses, and therefore requires some experience with such systems in order to improve interactions. This problem is re-examined here, with the aim of evaluat-ing how human users are in(cid:13)uenced by utterances built into a system based on the Wizard-of-Oz method. We report results which show that back-channel responses and brief con(cid:12)rmations from our system, have the e(cid:11)ects of prompting human spoken interactions and providing more human satisfaction.