Presentation 2008-12-10
Language Model Adaptation by Topic Model Based on Sequence of Words
Atsushi SAKO, Tetsuya TAKIGUCHI, Yasuo ARIKI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) It is important to consider semantics for reductions of recognition errors unlike humans or understanding meanings and contents. To accommodate these problems, Latent Semantic Analysis (LSA) or Probabilistic LSA have been proposed. However these methods are based on Bag-of-words techniques. For more sophisticated analysis, it needs to consider a sequence of words in a document. In this paper, we propose the method based on Kernel PCA and Dynamic Time Alignment Kernel in order to consider a sequence of words. Preliminary experimental results shows the proposed method can separete clearly a sequence of right turn/left turn prots data. Moreover, experimental results of language corpus shows the reduction of perplexity.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Latent Semantic Analysis / Kernel PCA / Topic Model
Paper # NLC2008-66,SP2008-121
Date of Issue

Conference Information
Committee NLC
Conference Date 2008/12/2(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) Language Model Adaptation by Topic Model Based on Sequence of Words
Sub Title (in English)
Keyword(1) Latent Semantic Analysis
Keyword(2) Kernel PCA
Keyword(3) Topic Model
1st Author's Name Atsushi SAKO
1st Author's Affiliation Guraduate School of Science and Technology, Kobe University:Guraduate School of Engineering, Kobe University()
2nd Author's Name Tetsuya TAKIGUCHI
2nd Author's Affiliation /
3rd Author's Name Yasuo ARIKI
3rd Author's Affiliation
Date 2008-12-10
Paper # NLC2008-66,SP2008-121
Volume (vol) vol.108
Number (no) 337
Page pp.pp.-
#Pages 6
Date of Issue