Estimation of pause length set by storyteller at sentence boundary in nursery tale: design of synthesized speech to support listener's comprehension

  • Authors:
  • Hideji Enokizu;Kazuhiro Uenosono;Seiichi Komiya

  • Affiliations:
  • Shibaura Institute of Technology, Tokyo, Japan;Graduate School of Engineering, Shibaura Institute of Technology, Tokyo, Japan;Graduate School of Engineering, Shibaura Institute of Technology, Tokyo, Japan

  • Venue:
  • WSEAS Transactions on Information Science and Applications
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

In the present study, we examined the pause length which the storyteller sets at each sentence boundary in the story text to support children's comprehension and proposed one method of estimating the pause length preliminarily to apply to the speech synthesis. While reading the story text, the reader constructs the situation model that is the mental microworld described in each sentence. Unlike the reader who can read the text at his own pace, however, the listener must hear the story at the storyteller's reading pace. Therefore we assumed that the experienced storyteller can set the pause length at each sentence boundary necessary for the listener to construct the situation model. If the listener is a young child, this assumption is true. Then we focused on the relationship between the mental operations for constructing the situation model and the pause length set by the experienced storyteller to help the listener comprehend the story. Specifically, we devised a knowledge representation with multilayered frame structure corresponding to the situation model to predict the mental operations for constructing the situation model. The knowledge representation was named as the situation frame. The results indicated that the storyteller controlled the pause length on the basis of several operations. The formula, that can estimate the pause length from the operation needed to construct the situation frame, seems to provide us the reasonable method to determine the pause length in the synthesized speech preliminarily.