Improving Naïve Bayes text classifier with modified EM algorithm

Han Joon Kim, Jae Young Chang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

This paper presents the method of significantly improving conventional Bayesian statistical text classifier by incorporating accelerated EM (Expectation Maximization) algorithm. EM algorithm experiences a slow convergence and performance degrade in its iterative process, especially when real textual documents do not follow EM’s assumptions. We propose a new accelerated EM algorithm that is simple yet has a fast convergence speed and allow to estimate a more accurate classification model on Bayesian text classifier.

Original languageEnglish
Title of host publicationFoundations of Intelligent Systems - 14th International Symposium, ISMIS 2003, Proceedings
EditorsNing Zhong, Zbigniew W. Ras, Shusaku Tsumoto, Einoshin Suzuki
PublisherSpringer Verlag
Pages326-333
Number of pages8
ISBN (Print)3540202560, 9783540202561
DOIs
StatePublished - 2003
Event14th International Symposium on Methodologies for Intelligent Systems, ISMIS 2003 - Maebashi City, Japan
Duration: 28 Oct 200331 Oct 2003

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2871
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference14th International Symposium on Methodologies for Intelligent Systems, ISMIS 2003
Country/TerritoryJapan
CityMaebashi City
Period28/10/0331/10/03

Keywords

  • Classification uncertainty
  • EM algorithm
  • Naïve Bayes
  • Selective sampling
  • Text classification

Fingerprint

Dive into the research topics of 'Improving Naïve Bayes text classifier with modified EM algorithm'. Together they form a unique fingerprint.

Cite this