Presentation 1995/7/21
Adaptive Scaling of a Semantic Space
Hideki Kozima, Akira Ito,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes a computationally feasible method for measuring context-sensitive semantic distance between words. The distance is computed by adaptive scaling of the semantic space where each word in the vocabulary V is represented by a multi-dimensional Q-vector. The vectors are obtained through a principal component analysis on the P-vectors generated from an English dictionary. Given a word set C which specified a context, each dimension of the semantic space is scaled up or down according to the distribution of C on the dimension. In the scaled space, word distance becomes sensitive to the context which comes out from C. An evaluation through word prediction shows that the proposed distance works well as an context-sensitive word distance.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) semantic distance / word similarity / association / thesauri / context processing
Paper #
Date of Issue

Conference Information
Committee NLC
Conference Date 1995/7/21(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) Adaptive Scaling of a Semantic Space
Sub Title (in English)
Keyword(1) semantic distance
Keyword(2) word similarity
Keyword(3) association
Keyword(4) thesauri
Keyword(5) context processing
1st Author's Name Hideki Kozima
1st Author's Affiliation Kansai Advanced Research Center Communications Research Laboratory()
2nd Author's Name Akira Ito
2nd Author's Affiliation Kansai Advanced Research Center Communications Research Laboratory
Date 1995/7/21
Paper #
Volume (vol) vol.95
Number (no) 169
Page pp.pp.-
#Pages 8
Date of Issue