Presentation	2003/11/14 Effects of Document Similarity Definitions on Classification Performance of Self-Organizing Maps Kazumi SAITO, Masumi ISHIKAWA,
PDF Download Page	PDF download Page Link
Abstract(in Japanese)	(See Japanese page)
Abstract(in English)	In this paper, we study effects on classification performance of self-organizing maps (SOM) caused by document similarity definitions, i.e., a standard Euclidean norm, logarithmic likelihood based on multinomial distributions, and cosine similarity. In terms of statistical model, we formalize three kinds of SOMs corresponding to these similarity definitions, and derive their learning algorithms. In the experiments using three kinds of benchmark document data sets, we evaluate these SOM models in the aspects of the classification performances and resulting self-organized maps.
Keyword(in Japanese)	(See Japanese page)
Keyword(in English)	self-organizing map / text mining / multinomial distribution / cosine similarity
Paper #	NC2003-73
Date of Issue

Conference Information
Committee	NC
Conference Date	2003/11/14(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	Neurocomputing (NC)
Language	JPN
Title (in Japanese)	(See Japanese page)
Sub Title (in Japanese)	(See Japanese page)
Title (in English)	Effects of Document Similarity Definitions on Classification Performance of Self-Organizing Maps
Sub Title (in English)
Keyword(1)	self-organizing map
Keyword(2)	text mining
Keyword(3)	multinomial distribution
Keyword(4)	cosine similarity
1st Author's Name	Kazumi SAITO
1st Author's Affiliation	NTT Communication Science Laboratories()
2nd Author's Name	Masumi ISHIKAWA
2nd Author's Affiliation	Kyushu Institute of Technology
Date	2003/11/14
Paper #	NC2003-73
Volume (vol)	vol.103
Number (no)	465
Page	pp.pp.-
#Pages	6
Date of Issue