Cue Phrase Selection in Instruction Dialogue Using Machine Learning

The purpose of this paper is to identify e ective factors for selecting discourse organization cue phrases in instruction dialogue that signal changes in discourse structure such as topic shifts and attentional state changes. By using a machine learning technique, a variety of features concerning discourse structure, task structure, and dialogue context are examined in terms of their e ectiveness and the best set of learning features is identi ed. Our result reveals that, in addition to discourse structure, already identi ed in previous studies, task structure and dialogue context play an important role. Moreover, an evaluation using a large dialogue corpus shows the utility of applying machine learning techniques to cue phrase selection.