Semi-supervised discourse relation classification with structural learning

Hugo Hernault, Danushka Bollegala, Mitsuru Ishizuka

研究成果: Conference contribution

9 被引用数 (Scopus)

抄録

The corpora available for training discourse relation classifiers are annotated using a general set of discourse relations. However, for certain applications, custom discourse relations are required. Creating a new annotated corpus with a new relation taxonomy is a time-consuming and costly process. We address this problem by proposing a semi-supervised approach to discourse relation classification based on Structural Learning. First, we solve a set of auxiliary classification problems using unlabeled data. Second, the learned classifiers are used to extend feature vectors to train a discourse relation classifier. By defining a relevant set of auxiliary classification problems, we show that the proposed method brings improvement of at least 50% in accuracy and F-score on the RST Discourse Treebank and Penn Discourse Treebank, when small training sets of ca. 1000 training instances are employed. This is an attractive perspective for training discourse relation classifiers on domains where little amount of labeled training data is available.

本文言語English
ホスト出版物のタイトルLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ページ340-352
ページ数13
6608 LNCS
PART 1
DOI
出版ステータスPublished - 2011
外部発表はい
イベント12th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2011 - Tokyo
継続期間: 2011 2 202011 2 26

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
番号PART 1
6608 LNCS
ISSN(印刷版)03029743
ISSN(電子版)16113349

Other

Other12th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2011
CityTokyo
Period11/2/2011/2/26

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

フィンガープリント 「Semi-supervised discourse relation classification with structural learning」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