Leveraging evaluation metric-related training criteria for speech summarization

Shih Hsiang Lin*, Yu Mei Chang, Jia Wen Liu, Berlin Chen

*此作品的通信作者

研究成果: 書貢獻/報告類型會議論文篇章

13 引文 斯高帕斯(Scopus)

摘要

Many of the existing machine-learning approaches to speech summarization cast important sentence selection as a two-class classification problem and have shown empirical success for a wide variety of summarization tasks. However, the imbalanced-data problem sometimes results in a trained speech summarizer with unsatisfactory performance. On the other hand, training the summarizer by improving the associated classification accuracy does not always lead to better summarization evaluation performance. In view of such phenomena, we hence investigate two different training criteria to alleviate the negative effects caused by them, as well as to boost the summarizer's performance. One is to learn the classification capability of a summarizer on the basis of the pair-wise ordering information of sentences in a training document according to a degree of importance. The other is to train the summarizer by directly maximizing the associated evaluation score. Experimental results on the broadcast news summarization task show that these two training criteria can give substantial improvements over the baseline SVM summarization system.

原文英語
主出版物標題2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010 - Proceedings
頁面5314-5317
頁數4
DOIs
出版狀態已發佈 - 2010
事件2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010 - Dallas, TX, 美国
持續時間: 2010 三月 142010 三月 19

出版系列

名字ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
ISSN(列印)1520-6149

其他

其他2010 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2010
國家/地區美国
城市Dallas, TX
期間2010/03/142010/03/19

ASJC Scopus subject areas

  • 軟體
  • 訊號處理
  • 電氣與電子工程

指紋

深入研究「Leveraging evaluation metric-related training criteria for speech summarization」主題。共同形成了獨特的指紋。

引用此