PP AMBIGUITY RESOLUTION THROUGH SEMANTIC MATCHING

スポンサーリンク

概要

• 論文の詳細を見る

• One of the crucial ambiguity problems in natural language processing is the prepositional phrase (PP) attachment resolution. In order to decide whether a prepositional phrase is adjectival or adverbial, we need to simultaneously employ different types of knowledge. This paper shows how a robust semantic hierarchy can facilitate the process of correct prepositional phrase attachment. We propose two semantic based methods both of which utilise such a semantic hierarchy. The first method is based on a calculation of the semantic distance between an unseen candidate for PP attachment and the matching example in the training corpus. The second method belongs to the family of inductive learning algorithms and derives a decision tree which then serves as an universal mechanism for the PP attachment resolution. We present the results comparing these methods with other corpus based approaches and discuss their feasibility in solving other NLP related problems.

• 一般社団法人情報処理学会の論文
• 1996-11-18

著者

• Nagao Makoto

  Department Of Electrical Engineering Ii Kyoto University

• Stetina Jiri

  Department of Electrical Engineering II Kyoto University

• Nagao Makoto

  Department Of Biological Sciences Tokyo Institute Of Technology

関連論文

• Grammar Writing System (GRADE) of Mu-Machine Translation Project and its Characteristics
• PP AMBIGUITY RESOLUTION THROUGH SEMANTIC MATCHING
• A System for the Analysis of Aerial Photographs and Their Preprocessing
• Analysis of Japanese Sentences by Using Semantic and Contextual Information (I)-Semantic Analysis
• PLATON : a New Programming Language for Natural Language Analysis
• Analysis of Japanese Sentences by Using Semantic and Contextual Information (II)-Contextual Analysis
• An Automatic Method of the Extraction of Important Words from Japanese Scientific Documents
• Frequency Coupling Model for Dynamics of Responses to Stimuli in Plasmodium of Physarum Polycephalum

スポンサーリンク