Semi-supervised learning for imbalanced sentiment classification

Shoushan Li; Zhongqing Wang; Guodong Zhou; Yat Mei Lee

doi:10.5591/978-1-57735-516-8/IJCAI11-306

Semi-supervised learning for imbalanced sentiment classification

Shoushan Li, Zhongqing Wang, Guodong Zhou, Yat Mei Lee

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

154 Citations (Scopus)

Abstract

Various semi-supervised learning methods have been proposed recently to solve the long-standing shortage problem of manually labeled data in sentiment classification. However, most existing studies assume the balance between negative and positive samples in both the labeled and unlabeled data, which may not be true in reality. In this paper, we investigate a more common case of semi-supervised learning for imbalanced sentiment classification. In particular, various random subspaces are dynamically generated to deal with the imbalanced class distribution problem. Evaluation across four domains shows the effectiveness of our approach.

Original language	English
Title of host publication	IJCAI 2011 - 22nd International Joint Conference on Artificial Intelligence
Pages	1826-1831
Number of pages	6
DOIs	https://doi.org/10.5591/978-1-57735-516-8/IJCAI11-306
Publication status	Published - 1 Dec 2011
Event	22nd International Joint Conference on Artificial Intelligence, IJCAI 2011 - Barcelona, Catalonia, Spain Duration: 16 Jul 2011 → 22 Jul 2011

Conference

Conference	22nd International Joint Conference on Artificial Intelligence, IJCAI 2011
Country/Territory	Spain
City	Barcelona, Catalonia
Period	16/07/11 → 22/07/11

ASJC Scopus subject areas

Artificial Intelligence

More information

10.5591/978-1-57735-516-8/IJCAI11-306

Cite this

@inproceedings{330cea18643d4a40b4fd5bfde443a5a7,

title = "Semi-supervised learning for imbalanced sentiment classification",

abstract = "Various semi-supervised learning methods have been proposed recently to solve the long-standing shortage problem of manually labeled data in sentiment classification. However, most existing studies assume the balance between negative and positive samples in both the labeled and unlabeled data, which may not be true in reality. In this paper, we investigate a more common case of semi-supervised learning for imbalanced sentiment classification. In particular, various random subspaces are dynamically generated to deal with the imbalanced class distribution problem. Evaluation across four domains shows the effectiveness of our approach.",

author = "Shoushan Li and Zhongqing Wang and Guodong Zhou and Lee, {Yat Mei}",

year = "2011",

month = dec,

day = "1",

doi = "10.5591/978-1-57735-516-8/IJCAI11-306",

language = "English",

isbn = "9781577355120",

pages = "1826--1831",

booktitle = "IJCAI 2011 - 22nd International Joint Conference on Artificial Intelligence",

note = "22nd International Joint Conference on Artificial Intelligence, IJCAI 2011 ; Conference date: 16-07-2011 Through 22-07-2011",

}

TY - GEN

T1 - Semi-supervised learning for imbalanced sentiment classification

AU - Li, Shoushan

AU - Wang, Zhongqing

AU - Zhou, Guodong

AU - Lee, Yat Mei

PY - 2011/12/1

Y1 - 2011/12/1

N2 - Various semi-supervised learning methods have been proposed recently to solve the long-standing shortage problem of manually labeled data in sentiment classification. However, most existing studies assume the balance between negative and positive samples in both the labeled and unlabeled data, which may not be true in reality. In this paper, we investigate a more common case of semi-supervised learning for imbalanced sentiment classification. In particular, various random subspaces are dynamically generated to deal with the imbalanced class distribution problem. Evaluation across four domains shows the effectiveness of our approach.

AB - Various semi-supervised learning methods have been proposed recently to solve the long-standing shortage problem of manually labeled data in sentiment classification. However, most existing studies assume the balance between negative and positive samples in both the labeled and unlabeled data, which may not be true in reality. In this paper, we investigate a more common case of semi-supervised learning for imbalanced sentiment classification. In particular, various random subspaces are dynamically generated to deal with the imbalanced class distribution problem. Evaluation across four domains shows the effectiveness of our approach.

UR - http://www.scopus.com/inward/record.url?scp=84881082798&partnerID=8YFLogxK

U2 - 10.5591/978-1-57735-516-8/IJCAI11-306

DO - 10.5591/978-1-57735-516-8/IJCAI11-306

M3 - Conference article published in proceeding or book

SN - 9781577355120

SP - 1826

EP - 1831

BT - IJCAI 2011 - 22nd International Joint Conference on Artificial Intelligence

T2 - 22nd International Joint Conference on Artificial Intelligence, IJCAI 2011

Y2 - 16 July 2011 through 22 July 2011

ER -

Semi-supervised learning for imbalanced sentiment classification

Abstract

Conference

ASJC Scopus subject areas

More information

Other files and links

Fingerprint

Cite this