Or gate Bayesian networks for text classification: A discriminative alternative approach to multinomial naive Bayes

Luis M. de Campos; J.M. Fernández-Luna; J.F. Huete; Alfonso E. Romero

Or gate Bayesian networks for text classification: A discriminative alternative approach to multinomial naive Bayes

Luis M. de Campos, J.M. Fernández-Luna, J.F. Huete, Alfonso E. Romero

Department of Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We propose a simple Bayesian network-based text classifier, which may be considered as a discriminative counterpart of the generative multinomial naive Bayes classifier. The method relies on the use of a fixed network topology with the arcs going form term nodes to class nodes, and also on a network parametrization based on noisy or gates. Comparative experiments of the proposed method with naive Bayes and Rocchio algorithms are carried out using three standard document collections.

Original language	English
Title of host publication	Actas del XIV Congreso Español sobre Tecnologías y Lógica Fuzzy
Pages	385-390
Number of pages	6
Publication status	Published - 2008
Event	XIV Congreso Español sobre Tecnologías y Lógica Fuzzy, ESTYLF 2008 - Mieres-Langreo, Spain Duration: 17 Sept 2008 → 19 Sept 2008

Conference

Conference	XIV Congreso Español sobre Tecnologías y Lógica Fuzzy, ESTYLF 2008
Country/Territory	Spain
City	Mieres-Langreo
Period	17/09/08 → 19/09/08

Keywords

Bayesian network
noisy or gate
multinomial naive Bayes
text classification

Access to Document

http://www.cs.rhul.ac.uk/~aeromero/pdf/estylf08-orgate.pdf

Cite this

@inproceedings{08cc33fe87c44077bfa16d4c58d960cd,

title = "Or gate Bayesian networks for text classification: A discriminative alternative approach to multinomial naive Bayes",

abstract = "We propose a simple Bayesian network-based text classifier, which may be considered as a discriminative counterpart of the generative multinomial naive Bayes classifier. The method relies on the use of a fixed network topology with the arcs going form term nodes to class nodes, and also on a network parametrization based on noisy or gates. Comparative experiments of the proposed method with naive Bayes and Rocchio algorithms are carried out using three standard document collections.",

keywords = "Bayesian network, noisy or gate, multinomial naive Bayes, text classification ",

author = "{de Campos}, {Luis M.} and J.M. Fern{\'a}ndez-Luna and J.F. Huete and Romero, {Alfonso E.}",

year = "2008",

language = "English",

isbn = "978-84-691-5807-4",

pages = "385--390",

booktitle = "Actas del XIV Congreso Espa{\~n}ol sobre Tecnolog{\'i}as y L{\'o}gica Fuzzy",

note = "XIV Congreso Espa{\~n}ol sobre Tecnolog{\'i}as y L{\'o}gica Fuzzy, ESTYLF 2008 ; Conference date: 17-09-2008 Through 19-09-2008",

}

de Campos, LM, Fernández-Luna, JM, Huete, JF & Romero, AE 2008, Or gate Bayesian networks for text classification: A discriminative alternative approach to multinomial naive Bayes. in Actas del XIV Congreso Español sobre Tecnologías y Lógica Fuzzy. pp. 385-390, XIV Congreso Español sobre Tecnologías y Lógica Fuzzy, ESTYLF 2008, Mieres-Langreo, Spain, 17/09/08. <http://www.cs.rhul.ac.uk/~aeromero/pdf/estylf08-orgate.pdf>

TY - GEN

T1 - Or gate Bayesian networks for text classification: A discriminative alternative approach to multinomial naive Bayes

AU - de Campos, Luis M.

AU - Fernández-Luna, J.M.

AU - Huete, J.F.

AU - Romero, Alfonso E.

PY - 2008

Y1 - 2008

N2 - We propose a simple Bayesian network-based text classifier, which may be considered as a discriminative counterpart of the generative multinomial naive Bayes classifier. The method relies on the use of a fixed network topology with the arcs going form term nodes to class nodes, and also on a network parametrization based on noisy or gates. Comparative experiments of the proposed method with naive Bayes and Rocchio algorithms are carried out using three standard document collections.

AB - We propose a simple Bayesian network-based text classifier, which may be considered as a discriminative counterpart of the generative multinomial naive Bayes classifier. The method relies on the use of a fixed network topology with the arcs going form term nodes to class nodes, and also on a network parametrization based on noisy or gates. Comparative experiments of the proposed method with naive Bayes and Rocchio algorithms are carried out using three standard document collections.

KW - Bayesian network

KW - noisy or gate

KW - multinomial naive Bayes

KW - text classification

M3 - Conference contribution

SN - 978-84-691-5807-4

SP - 385

EP - 390

BT - Actas del XIV Congreso Español sobre Tecnologías y Lógica Fuzzy

T2 - XIV Congreso Español sobre Tecnologías y Lógica Fuzzy, ESTYLF 2008

Y2 - 17 September 2008 through 19 September 2008

ER -