Presentation 2009-03-13
Word Association based on α-divergence and Application to Document Classifcation
Katsuji BESSHO, Toshio UCHIYAMA, Tadasu Uchiyama,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The co-occurrence vectors that are the co-occurrence frequencies between words and semantic attributes in a corpus reflect the semantic similarities between words. This paper shows that words of superordinate or subordinate or sibling concept at various levels are associated by applying α-divergence that is the distance measure to which Kullback-Leibler Metric is expanded as the distance measure between co-occurrence vectors. Moreover, it describes the application of α-divergence between co-occurrence vectors to the document classification.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Co-occurrence Vector / α-Divergence / Word Association / Document Classification
Paper # PRMU2008-255
Date of Issue

Conference Information
Committee PRMU
Conference Date 2009/3/6(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 Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Word Association based on α-divergence and Application to Document Classifcation
Sub Title (in English)
Keyword(1) Co-occurrence Vector
Keyword(2) α-Divergence
Keyword(3) Word Association
Keyword(4) Document Classification
1st Author's Name Katsuji BESSHO
1st Author's Affiliation NTT Cyber Solutions Laboratories, NTT Corporation()
2nd Author's Name Toshio UCHIYAMA
2nd Author's Affiliation NTT Cyber Solutions Laboratories, NTT Corporation
3rd Author's Name Tadasu Uchiyama
3rd Author's Affiliation NTT Cyber Solutions Laboratories, NTT Corporation
Date 2009-03-13
Paper # PRMU2008-255
Volume (vol) vol.108
Number (no) 484
Page pp.pp.-
#Pages 6
Date of Issue