Presentation 2000/10/20
A Method for Resolving Overlapping Ambiguities in Chinese Word Segmentation Process
Dongli Han, Haodong Wu, Teiji Furugori,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We propose an efficient method of selecting correct word sequence by resolving overlapping ambiguities found in sentencial analyses of chinese Language. The method works in the following manner. We detect the overlapping ambiguities by a FBMM scanner, and then, using Internet data, the ambiguities are resolved by a statistical measure called the relevancy value(RV)that is to determine the likelihood of occurrence for each ambiguous word sequence. We conducted an experiment and got correct word sequences at a success rate of 84.1%.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Chinese / word segmentation / overlapping ambiguity / FBMM / Internet Corpus / relevancy value
Paper # NLC2000-25
Date of Issue

Conference Information
Committee NLC
Conference Date 2000/10/20(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 ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Method for Resolving Overlapping Ambiguities in Chinese Word Segmentation Process
Sub Title (in English)
Keyword(1) Chinese
Keyword(2) word segmentation
Keyword(3) overlapping ambiguity
Keyword(4) FBMM
Keyword(5) Internet Corpus
Keyword(6) relevancy value
1st Author's Name Dongli Han
1st Author's Affiliation Department of Computer Science, the University of Electro-Communications()
2nd Author's Name Haodong Wu
2nd Author's Affiliation Department of Language and Culture, Dokkyo University
3rd Author's Name Teiji Furugori
3rd Author's Affiliation Department of Computer Science, the University of Electro-Communications
Date 2000/10/20
Paper # NLC2000-25
Volume (vol) vol.100
Number (no) 401
Page pp.pp.-
#Pages 6
Date of Issue