Fast dimension reduction for document classification based on Imprecise Spectrum Analysis

Hu Guan; Bin Xiao; Jingyu Zhou; Minyi Guo; Tao Yang

doi:10.1145/1871437.1871721

Fast dimension reduction for document classification based on Imprecise Spectrum Analysis

Hu Guan, Bin Xiao, Jingyu Zhou, Minyi Guo, Tao Yang

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

3 Citations (Scopus)

Abstract

This paper proposes an algorithm called Imprecise Spectrum Analysis (ISA) to carry out fast dimension reduction for document classification. ISA is designed based on the one-sided Jacobi method for Singular Value Decomposition (SVD). To speedup dimension reduction, it simplifies the orthog-onalization process of Jacobi computation and introduces a new mapping formula for transforming original DOCument-term vectors. To improve classification accuracy using ISA, a feature selection method is further developed to make inter-class feature vectors more orthogonal in building the initial weighted term-document matrix. Our experimental results show that ISA is extremely fast in handling large term-document matrices and delivers better or competitive classification accuracy compared to SVD-based LSI.

Original language	English
Title of host publication	CIKM'10 - Proceedings of the 19th International Conference on Information and Knowledge Management and Co-located Workshops
Pages	1753-1756
Number of pages	4
DOIs	https://doi.org/10.1145/1871437.1871721
Publication status	Published - 1 Dec 2010
Event	19th International Conference on Information and Knowledge Management and Co-located Workshops, CIKM'10 - Toronto, ON, Canada Duration: 26 Oct 2010 → 30 Oct 2010

Conference

Conference	19th International Conference on Information and Knowledge Management and Co-located Workshops, CIKM'10
Country/Territory	Canada
City	Toronto, ON
Period	26/10/10 → 30/10/10

Keywords

Dimension reduction
Feature selection
LSI
SVD

ASJC Scopus subject areas

General Decision Sciences
General Business,Management and Accounting

More information

10.1145/1871437.1871721

Cite this

@inproceedings{a80312eb4ccb45ddbdd857393f09eb76,

title = "Fast dimension reduction for document classification based on Imprecise Spectrum Analysis",

abstract = "This paper proposes an algorithm called Imprecise Spectrum Analysis (ISA) to carry out fast dimension reduction for document classification. ISA is designed based on the one-sided Jacobi method for Singular Value Decomposition (SVD). To speedup dimension reduction, it simplifies the orthog-onalization process of Jacobi computation and introduces a new mapping formula for transforming original DOCument-term vectors. To improve classification accuracy using ISA, a feature selection method is further developed to make inter-class feature vectors more orthogonal in building the initial weighted term-document matrix. Our experimental results show that ISA is extremely fast in handling large term-document matrices and delivers better or competitive classification accuracy compared to SVD-based LSI.",

keywords = "Dimension reduction, Feature selection, LSI, SVD",

author = "Hu Guan and Bin Xiao and Jingyu Zhou and Minyi Guo and Tao Yang",

year = "2010",

month = dec,

day = "1",

doi = "10.1145/1871437.1871721",

language = "English",

isbn = "9781450300995",

pages = "1753--1756",

booktitle = "CIKM'10 - Proceedings of the 19th International Conference on Information and Knowledge Management and Co-located Workshops",

note = "19th International Conference on Information and Knowledge Management and Co-located Workshops, CIKM'10 ; Conference date: 26-10-2010 Through 30-10-2010",

}

Guan, H, Xiao, B, Zhou, J, Guo, M & Yang, T 2010, Fast dimension reduction for document classification based on Imprecise Spectrum Analysis. in CIKM'10 - Proceedings of the 19th International Conference on Information and Knowledge Management and Co-located Workshops. pp. 1753-1756, 19th International Conference on Information and Knowledge Management and Co-located Workshops, CIKM'10, Toronto, ON, Canada, 26/10/10. https://doi.org/10.1145/1871437.1871721

Fast dimension reduction for document classification based on Imprecise Spectrum Analysis. / Guan, Hu; Xiao, Bin; Zhou, Jingyu et al.
CIKM'10 - Proceedings of the 19th International Conference on Information and Knowledge Management and Co-located Workshops. 2010. p. 1753-1756.

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - Fast dimension reduction for document classification based on Imprecise Spectrum Analysis

AU - Guan, Hu

AU - Xiao, Bin

AU - Zhou, Jingyu

AU - Guo, Minyi

AU - Yang, Tao

PY - 2010/12/1

Y1 - 2010/12/1

N2 - This paper proposes an algorithm called Imprecise Spectrum Analysis (ISA) to carry out fast dimension reduction for document classification. ISA is designed based on the one-sided Jacobi method for Singular Value Decomposition (SVD). To speedup dimension reduction, it simplifies the orthog-onalization process of Jacobi computation and introduces a new mapping formula for transforming original DOCument-term vectors. To improve classification accuracy using ISA, a feature selection method is further developed to make inter-class feature vectors more orthogonal in building the initial weighted term-document matrix. Our experimental results show that ISA is extremely fast in handling large term-document matrices and delivers better or competitive classification accuracy compared to SVD-based LSI.

AB - This paper proposes an algorithm called Imprecise Spectrum Analysis (ISA) to carry out fast dimension reduction for document classification. ISA is designed based on the one-sided Jacobi method for Singular Value Decomposition (SVD). To speedup dimension reduction, it simplifies the orthog-onalization process of Jacobi computation and introduces a new mapping formula for transforming original DOCument-term vectors. To improve classification accuracy using ISA, a feature selection method is further developed to make inter-class feature vectors more orthogonal in building the initial weighted term-document matrix. Our experimental results show that ISA is extremely fast in handling large term-document matrices and delivers better or competitive classification accuracy compared to SVD-based LSI.

KW - Dimension reduction

KW - Feature selection

KW - LSI

KW - SVD

UR - http://www.scopus.com/inward/record.url?scp=78651295006&partnerID=8YFLogxK

U2 - 10.1145/1871437.1871721

DO - 10.1145/1871437.1871721

M3 - Conference article published in proceeding or book

SN - 9781450300995

SP - 1753

EP - 1756

BT - CIKM'10 - Proceedings of the 19th International Conference on Information and Knowledge Management and Co-located Workshops

T2 - 19th International Conference on Information and Knowledge Management and Co-located Workshops, CIKM'10

Y2 - 26 October 2010 through 30 October 2010

ER -

Fast dimension reduction for document classification based on Imprecise Spectrum Analysis

Abstract

Conference

Keywords

ASJC Scopus subject areas

More information

Other files and links

Fingerprint

Cite this