A new method of parameter estimation for multinomial naive Bayes text classifiers

Sang Bum Kim; Hae Chang Rim; Heui Seok Lim

doi:10.1145/564376.564459

A new method of parameter estimation for multinomial naive Bayes text classifiers

Sang Bum Kim, Hae Chang Rim, Heui Seok Lim

Research output: Contribution to journal › Conference article › peer-review

7 Citations (Scopus)

Abstract

Multinomial naive Bayes classifies have been widely used for the probabilistic text classification. However, their parameter estimation method sometimes generates inappropriate probabilities. In this paper, we propose a topic document model approach for naive Bayes text classification, where their parameters are estimated with an expectation from the training documents. Experiments are conducted on Reuters 21578 and 20 Newsgroup collection, and our proposed approach obtained a significant improvement in performance over the conventional approach.

Original language	English
Pages (from-to)	391-392
Number of pages	2
Journal	SIGIR Forum (ACM Special Interest Group on Information Retrieval)
DOIs	https://doi.org/10.1145/564376.564459
Publication status	Published - 2002
Event	Proceedings of the Twenty-Fifth Annual International ACM SIGIR Conference on Research and Development in Information Retrieval - Tampere, Finland Duration: 2002 Aug 11 → 2002 Aug 15

ASJC Scopus subject areas

Management Information Systems
Hardware and Architecture

Access to Document

10.1145/564376.564459

Cite this

@article{8dae4f3f954b42c7b91ea9064ac8036c,

title = "A new method of parameter estimation for multinomial naive Bayes text classifiers",

abstract = "Multinomial naive Bayes classifies have been widely used for the probabilistic text classification. However, their parameter estimation method sometimes generates inappropriate probabilities. In this paper, we propose a topic document model approach for naive Bayes text classification, where their parameters are estimated with an expectation from the training documents. Experiments are conducted on Reuters 21578 and 20 Newsgroup collection, and our proposed approach obtained a significant improvement in performance over the conventional approach.",

author = "Kim, {Sang Bum} and Rim, {Hae Chang} and Lim, {Heui Seok}",

year = "2002",

doi = "10.1145/564376.564459",

language = "English",

pages = "391--392",

journal = "SIGIR Forum (ACM Special Interest Group on Information Retrieval)",

issn = "0163-5840",

publisher = "Association for Computing Machinery (ACM)",

note = "Proceedings of the Twenty-Fifth Annual International ACM SIGIR Conference on Research and Development in Information Retrieval ; Conference date: 11-08-2002 Through 15-08-2002",

}

TY - JOUR

T1 - A new method of parameter estimation for multinomial naive Bayes text classifiers

AU - Kim, Sang Bum

AU - Rim, Hae Chang

AU - Lim, Heui Seok

PY - 2002

Y1 - 2002

N2 - Multinomial naive Bayes classifies have been widely used for the probabilistic text classification. However, their parameter estimation method sometimes generates inappropriate probabilities. In this paper, we propose a topic document model approach for naive Bayes text classification, where their parameters are estimated with an expectation from the training documents. Experiments are conducted on Reuters 21578 and 20 Newsgroup collection, and our proposed approach obtained a significant improvement in performance over the conventional approach.

AB - Multinomial naive Bayes classifies have been widely used for the probabilistic text classification. However, their parameter estimation method sometimes generates inappropriate probabilities. In this paper, we propose a topic document model approach for naive Bayes text classification, where their parameters are estimated with an expectation from the training documents. Experiments are conducted on Reuters 21578 and 20 Newsgroup collection, and our proposed approach obtained a significant improvement in performance over the conventional approach.

UR - http://www.scopus.com/inward/record.url?scp=0036993193&partnerID=8YFLogxK

U2 - 10.1145/564376.564459

DO - 10.1145/564376.564459

M3 - Conference article

AN - SCOPUS:0036993193

SN - 0163-5840

SP - 391

EP - 392

JO - SIGIR Forum (ACM Special Interest Group on Information Retrieval)

JF - SIGIR Forum (ACM Special Interest Group on Information Retrieval)

T2 - Proceedings of the Twenty-Fifth Annual International ACM SIGIR Conference on Research and Development in Information Retrieval

Y2 - 11 August 2002 through 15 August 2002

ER -

A new method of parameter estimation for multinomial naive Bayes text classifiers

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this