Effective Spoken Natural Language Dialog Requires Variable Initiative Behavior: An Empirical Study

Contents

The usefulness of spoken natural language dialog for human-computer interaction is unquestioned. A spoken natural language interface provides a flexible means of communication while enabling human users to keep their hands and eyes busy on the task at hand. In this paper we claim that spoken natural language dialog is even more useful if the participating system permits variable initiative behavior. This is behavior where the task initiative can vary from strongly computer controlled to strongly user controlled or somewhere in between. Such behavior allows a system to effectively communicate with both task novices and experts as well as users with intermediate levels of expertise. In support of this claim, we present experimental results from a study of an implemented system that can operate at differing levels of initiative while engaging in spoken natural language dialog with users in order to provide assistance in repairing an electronic circuit. The results are based on eight users who participated in a total of 141 dialogs speaking 2840 utterances.