Reducing lexical features in parsing by word embeddings

Hiroya Komatsu, Zen Den, Naoaki Okazaki, Kentaro Inui

研究成果: Paper査読

2 被引用数 (Scopus)

抄録

The high-dimensionality of lexical features in parsing can be memory consuming and cause over-fitting problems. We propose a general framework to replace all lexical feature templates by low-dimensional features induced from word embeddings. Applied to a near state-of-the-art dependency parser (Huang et al., 2012), our method improves the baseline, performs better than using cluster bit string features, and outperforms a recent neural network based parser. A further analysis shows that our framework has the effect hypothesized by Andreas and Klein (2014), namely (i) connecting unseen words to known ones, and (ii) encouraging common behaviors among invocabulary words.

本文言語English
ページ106-113
ページ数8
出版ステータスPublished - 2015 1月 1
イベント29th Pacific Asia Conference on Language, Information and Computation, PACLIC 2015 - Shanghai, China
継続期間: 2015 10月 302015 11月 1

Other

Other29th Pacific Asia Conference on Language, Information and Computation, PACLIC 2015
国/地域China
CityShanghai
Period15/10/3015/11/1

ASJC Scopus subject areas

  • 人工知能
  • 人間とコンピュータの相互作用
  • 言語学および言語

フィンガープリント

「Reducing lexical features in parsing by word embeddings」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