Prepositional Phrase Attachment Disambiguation Using Preference Rules and Conceptual Information
スポンサーリンク
概要
- 論文の詳細を見る
The disambiguation of prepositional phrase(PP) modifiers in unrestricted text is a difficult problem to resolve in NLP. In this paper, we describe an method which use preference rule and conceptual information from a machine-readable dictionary to disambiguate PP attachment. In case of an unique attachment can't be determined, probability from an annotated corpus is calculated to determine the attachment. As other work do, we use four head words of verb(v), head noun(n1) ahead of preposition, preposition(p) and head noun(n2) of the object of preposition to simplify the context we use in disambiguation.
- 一般社団法人情報処理学会の論文
- 1996-03-06
著者
関連論文
- 単語の結束性にもとづいてテキストを場面に分割する試み
- 英語辞書を利用した単語の類似度の計算
- 日本人の英語の特性を考慮した誤綴訂正
- インタネット上のデータを利用しての中国語単語分割における曖昧性の解消
- 概念情報に基づく前置詞句係り先の曖昧性の解消
- Prepositional Phrase Attachment Disambiguation Using Preference Rules and Conceptual Information
- Right Attachment of Prepositional Phrases by Using Reference Strategy
- Natural Language Proeessing for MPEG-7 Applications
- Adding Co-occurence Information to Dependency Grammar
- 係り受け関係と相互情報量を用いた単語の意味獲得