摘要
Extractive speech summarization attempts to select a representative set of sentences from a spoken document so as to succinctly describe the main theme of the original document. In this paper, we adapt the notion of risk minimization for extractive speech summarization by formulating the selection of summary sentences as a decision-making problem. To this end, we develop several selection strategies and modeling paradigms that can leverage supervised and unsupervised summarization models to inherit their individual merits as well as to overcome their inherent limitations. On top of that, various component models are introduced, providing a principled way to render the redundancy and coherence relationships among sentences and between sentences and the whole document, respectively. A series of experiments on speech summarization seem to demonstrate that the methods deduced from our summarization framework are very competitive with existing summarization methods.
| 原文 | 英語 |
|---|---|
| 文章編號 | 5876303 |
| 頁（從 - 到） | 199-210 |
| 頁數 | 12 |
| 期刊 | IEEE Transactions on Audio, Speech and Language Processing |
| 卷 | 20 |
| 發行號 | 1 |
| DOIs | |
| 出版狀態 | 已發佈 - 2012 一月 1 |
指紋
ASJC Scopus subject areas
- Acoustics and Ultrasonics
- Electrical and Electronic Engineering
引用此文
A risk-aware modeling framework for speech summarization. / Chen, Berlin; Lin, Shih Hsiang.
於: IEEE Transactions on Audio, Speech and Language Processing, 卷 20, 編號 1, 5876303, 01.01.2012, p. 199-210.研究成果: 雜誌貢獻 › 文章
}
TY - JOUR
T1 - A risk-aware modeling framework for speech summarization
AU - Chen, Berlin
AU - Lin, Shih Hsiang
PY - 2012/1/1
Y1 - 2012/1/1
N2 - Extractive speech summarization attempts to select a representative set of sentences from a spoken document so as to succinctly describe the main theme of the original document. In this paper, we adapt the notion of risk minimization for extractive speech summarization by formulating the selection of summary sentences as a decision-making problem. To this end, we develop several selection strategies and modeling paradigms that can leverage supervised and unsupervised summarization models to inherit their individual merits as well as to overcome their inherent limitations. On top of that, various component models are introduced, providing a principled way to render the redundancy and coherence relationships among sentences and between sentences and the whole document, respectively. A series of experiments on speech summarization seem to demonstrate that the methods deduced from our summarization framework are very competitive with existing summarization methods.
AB - Extractive speech summarization attempts to select a representative set of sentences from a spoken document so as to succinctly describe the main theme of the original document. In this paper, we adapt the notion of risk minimization for extractive speech summarization by formulating the selection of summary sentences as a decision-making problem. To this end, we develop several selection strategies and modeling paradigms that can leverage supervised and unsupervised summarization models to inherit their individual merits as well as to overcome their inherent limitations. On top of that, various component models are introduced, providing a principled way to render the redundancy and coherence relationships among sentences and between sentences and the whole document, respectively. A series of experiments on speech summarization seem to demonstrate that the methods deduced from our summarization framework are very competitive with existing summarization methods.
KW - Decision-making
KW - language modeling
KW - loss functions
KW - risk minimization
KW - speech summarization
UR - http://www.scopus.com/inward/record.url?scp=81155132373&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=81155132373&partnerID=8YFLogxK
U2 - 10.1109/TASL.2011.2159596
DO - 10.1109/TASL.2011.2159596
M3 - Article
AN - SCOPUS:81155132373
VL - 20
SP - 199
EP - 210
JO - IEEE Transactions on Audio, Speech and Language Processing
JF - IEEE Transactions on Audio, Speech and Language Processing
SN - 1558-7916
IS - 1
M1 - 5876303
ER -