The annotation of Information Structure in spoken Japanese

This paper presents the main results of a pilot aimed at verifying the consistency of the Language into Act Theory model for the annotation of Information Structure in spoken Japanese. The segmentation of the Japanese speech flow into utterances through the detection of terminal prosodic breaks and the segmentation into information unit through non-terminal breaks works fine in Japanese. The main Information unit types characterizing the L-AcT approach (Comment, Topic, Parenthesis, Appendix) and their main properties, fit well with the Japanese data-set.