Simulating Spoken Dialogue with a Focus on Realistic Turn-taking
نویسنده
چکیده
We present a system for testing turn-taking strategies in a simulation environment, in which artificial dialogue participants exchange audio streams in real time – unlike earlier turn-taking simulations, which interchanged unambiguous symbolic messages. Dialogue participants autonomously determine their turn-taking behaviour, based on their analysis of the incoming audio. We use machine-learning methods to classifiy the continuous audio signal into symbolic turn-taking states. We experiment with various rule sets and show how simple, local management rules can create realistic behavioural patterns.
منابع مشابه
On-Line Learning of a Persian Spoken Dialogue System Using Real Training Data
The first spoken dialogue system developed for the Persian language is introduced. This is a ticket reservation system with Persian ASR and NLU modules. The focus of the paper is on learning the dialogue management module. In this work, real on-line training data are used during the learning process. For on-line learning, the effect of the variations of discount factor (g) on the learning speed...
متن کاملOn-Line Learning of a Persian Spoken Dialogue System Using Real Training Data
The first spoken dialogue system developed for the Persian language is introduced. This is a ticket reservation system with Persian ASR and NLU modules. The focus of the paper is on learning the dialogue management module. In this work, real on-line training data are used during the learning process. For on-line learning, the effect of the variations of discount factor (g) on the learning speed...
متن کاملA Bidding Approach to Turn-Taking
We propose a new turn-taking framework for spoken dialogue systems in which conversants bid for the turn. This differs from most current turn-taking approaches, where the turn only changes after the holder has released it. Our new framework uses Reinforcement Learning to choose appropriate turn bids, which indirectly estimates the utterance importance. We evaluate this approach in an artificial...
متن کاملThe Effect of CMC in Business Emails in Lingua Franca: Discourse Features and Misunderstandings
The paper argues that everyday exchange of business emails produces a development in the work-group relationship, which, in turn, makes new communication styles possible and acceptable by the users' habit to computer-mediated forms, even in unbalanced professional exchanges. The focus is on the (spoken) discourse features of email messages in a self-compiled corpus of selected computer-mediated...
متن کاملNon-humanlike Spoken Dialogue: A Design Perspective
We propose a non-humanlike spoken dialogue design, which consists of two elements: non-humanlike turn-taking and non-humanlike acknowledgment. Two experimental studies are reported in this paper. The first study shows that the proposed non-humanlike spoken dialogue design is effective for reducing speech collisions. It also presents pieces of evidence that show quick humanlike turn-taking is le...
متن کامل