Task adaptation using map estimation in N-gram language modeling

Hirokazu Masataki*, Yoshinori Sagisaka, Kazuya Hisaki, Tatsuya Kawahara

*この研究の対応する著者

研究成果: Conference article査読

24 被引用数 (Scopus)

抄録

This paper describes a method of task adaptation in N-gram language modeling, for accurately estimating the N-gram statistics from the small amount of data of the target task. Assuming a task-independent N-gram to be a-priori knowledge, the N-gram is adapted to a target task by MAP (maximum a-posteriori probability) estimation. Experimental results showed that the perplexities of the task adapted models were 15% (trigram), 24% (bigram) lower than those of the task-independent model, and that the perplexity reduction of the adaptation went up to 39% at maximum when the amount of text data in the adapted task was very small.

本文言語English
ページ(範囲)783-786
ページ数4
ジャーナルICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
2
出版ステータスPublished - 1997 1月 1
外部発表はい
イベントProceedings of the 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP. Part 1 (of 5) - Munich, Ger
継続期間: 1997 4月 211997 4月 24

ASJC Scopus subject areas

  • ソフトウェア
  • 信号処理
  • 電子工学および電気工学

フィンガープリント

「Task adaptation using map estimation in N-gram language modeling」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル