On Combining Classifier Mass Functions for Text Categorization

David A. Bell, Ji-wen W. Guan, Yaxin Bi

Research output: Contribution to journalArticle

32 Citations (Scopus)

Abstract

Experience shows that different text classification methods can give different results. We look here at a way of combining the results of two or more different classification methods using an evidential approach. The specific methods we have been experimenting with in our group include the support vector machine, kNN (nearest neighbors), kNN model-based approach (kNNM), and Rocchio methods, but the analysis and methods apply to any methods. We review these learning methods briefly, and then we describe our method for combining the classifiers. In a previous study, we suggested that the combination could be done using evidential operations and that using only two focal points in the mass functions gives good results. However, there are conditions under which we should choose to use more focal points. We assess some aspects of this choice from an reasoning perspective and suggest a refinement of the approach.
LanguageEnglish
Pages1307-1319
JournalKnowledge and Data Engineering, IEEE Transactions on
Volume17
Issue number10
DOIs
Publication statusPublished - 2005

Fingerprint

method
learning
support vector machine
analysis

Cite this

@article{0437577f17824a66a8d545b3f351f493,
title = "On Combining Classifier Mass Functions for Text Categorization",
abstract = "Experience shows that different text classification methods can give different results. We look here at a way of combining the results of two or more different classification methods using an evidential approach. The specific methods we have been experimenting with in our group include the support vector machine, kNN (nearest neighbors), kNN model-based approach (kNNM), and Rocchio methods, but the analysis and methods apply to any methods. We review these learning methods briefly, and then we describe our method for combining the classifiers. In a previous study, we suggested that the combination could be done using evidential operations and that using only two focal points in the mass functions gives good results. However, there are conditions under which we should choose to use more focal points. We assess some aspects of this choice from an reasoning perspective and suggest a refinement of the approach.",
author = "Bell, {David A.} and Guan, {Ji-wen W.} and Yaxin Bi",
year = "2005",
doi = "10.1109/TKDE.2005.167",
language = "English",
volume = "17",
pages = "1307--1319",
journal = "IEEE Transactions on Knowledge and Data Engineering",
issn = "1041-4347",
number = "10",

}

On Combining Classifier Mass Functions for Text Categorization. / Bell, David A.; Guan, Ji-wen W.; Bi, Yaxin.

In: Knowledge and Data Engineering, IEEE Transactions on, Vol. 17, No. 10, 2005, p. 1307-1319.

Research output: Contribution to journalArticle

TY - JOUR

T1 - On Combining Classifier Mass Functions for Text Categorization

AU - Bell, David A.

AU - Guan, Ji-wen W.

AU - Bi, Yaxin

PY - 2005

Y1 - 2005

N2 - Experience shows that different text classification methods can give different results. We look here at a way of combining the results of two or more different classification methods using an evidential approach. The specific methods we have been experimenting with in our group include the support vector machine, kNN (nearest neighbors), kNN model-based approach (kNNM), and Rocchio methods, but the analysis and methods apply to any methods. We review these learning methods briefly, and then we describe our method for combining the classifiers. In a previous study, we suggested that the combination could be done using evidential operations and that using only two focal points in the mass functions gives good results. However, there are conditions under which we should choose to use more focal points. We assess some aspects of this choice from an reasoning perspective and suggest a refinement of the approach.

AB - Experience shows that different text classification methods can give different results. We look here at a way of combining the results of two or more different classification methods using an evidential approach. The specific methods we have been experimenting with in our group include the support vector machine, kNN (nearest neighbors), kNN model-based approach (kNNM), and Rocchio methods, but the analysis and methods apply to any methods. We review these learning methods briefly, and then we describe our method for combining the classifiers. In a previous study, we suggested that the combination could be done using evidential operations and that using only two focal points in the mass functions gives good results. However, there are conditions under which we should choose to use more focal points. We assess some aspects of this choice from an reasoning perspective and suggest a refinement of the approach.

U2 - 10.1109/TKDE.2005.167

DO - 10.1109/TKDE.2005.167

M3 - Article

VL - 17

SP - 1307

EP - 1319

JO - IEEE Transactions on Knowledge and Data Engineering

T2 - IEEE Transactions on Knowledge and Data Engineering

JF - IEEE Transactions on Knowledge and Data Engineering

SN - 1041-4347

IS - 10

ER -