Presentation 1998/7/23
Post-processing of Japanese Morphological Analysis Using Transformation Rules and Contextual Information
Toru Hisamitsu, Yoshiki Niwa,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A method is proposed for the post-processing of Japanese morphological analysis using transformation rules and contextual information. The method corrects both segmentation errors and part-of-speech tagging errors. The transformation rules are acquired automatically by error-driven supervised learning. The rules consist of various types, such as lexicalized rules and schematic rules. Each rule is assigned a value for reliability. The rules are not specifically tailored for detecting unregistered words, but can correct errors caused by unregistered words. In addition, we propose the use of contextual information obtained from the result of analysis of neighboring sentences. The information reinforces unregistered word detection and disambiguation. The post-processing improved the precision of the analysis of an open corpus by 3%.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) morphological analysis / post-processing / error-driven supervised learning
Paper # NLC98-14
Date of Issue

Conference Information
Committee NLC
Conference Date 1998/7/23(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Post-processing of Japanese Morphological Analysis Using Transformation Rules and Contextual Information
Sub Title (in English)
Keyword(1) morphological analysis
Keyword(2) post-processing
Keyword(3) error-driven supervised learning
1st Author's Name Toru Hisamitsu
1st Author's Affiliation Advanced Research Laboratory, Hitachi, Ltd.()
2nd Author's Name Yoshiki Niwa
2nd Author's Affiliation Advanced Research Laboratory, Hitachi, Ltd.
Date 1998/7/23
Paper # NLC98-14
Volume (vol) vol.98
Number (no) 209
Page pp.pp.-
#Pages 8
Date of Issue