Evaluation and Advice System for Improving the Manner of Speaking in Lectures Using Features of Filled Pauses
This paper describes an evaluation and advice system for improving the manner of speaking in lectures. Our developed system evaluates the manner of speaking using decision trees. These decision trees are trained using features of .lled pauses that are included in lecture speeches. Recently, it has been learned that .lled pauses in speech in.uence the understanding of listeners and change the overall impression of the speech. Our decision-tree-based evaluation and advice system detects .lled pauses using automatic speech recognition. The system then uses the features of the .lled pauses in decision trees. Finally, the system evaluates the manner of speaking at three evaluation levels and provides advice to the speaker. In the evaluation experiment, our system worked well and advice given improved the manner of speaking in four out of .ve instances.
Wataru Naito Hiromitsu Nishizaki Yoshihiro Sekiguchi
Department of Education Interdisciplinary Graduate School of Medicine and Engineering, University of Department of Research Interdisciplinary Graduate School of Medicine and Engineering, University of
国际会议
2011亚太信号与信息处理协会年度峰会(APSIPAASC 2011)
西安
英文
1-4
2011-10-18（万方平台首次上网日期，不代表论文的发表时间）