Learning decision trees to determine turn-taking by spoken dialogue systems

Thispaper presents amethod for deciding the timing ofturn-taking in spoken dialogue systems. This method uses a decision tree learned from the corpus of dialogues between human users and systems in which desirable turn-taking behaviors are annotated by hand. It utilizes a variety of attributes, such as recognition and un- derstanding results and prosodic information. Unlike most of the existing systems it enables spoken dialogue systems to decide the timing of turn-taking based on not only pauses but also other fea- tures, so that users can speak to the system even if they put pauses in the middle of their utterances. The result of a preliminary exper- iment shows that the learned decision tree outperforms the baseline strategy, which takes turn at every user pauses.