Atlantis Press Paper Details: Sentiment Classification for Consumer Word-of-Mouth in Chinese: Comparison between Supervised and Unsupervised Approaches

title:

Sentiment Classification for Consumer Word-of-Mouth in Chinese: Comparison between Supervised and Unsupervised Approaches

publication:

ICEBI-10

part of series:

Advances in Intelligent Systems Research

ISBN:

978-90-78677-40-6

ISSN:

1951-6851

DOI:

doi:10.2991/icebi.2010.56 (how to use a DOI)

author(s):

Ziqing Zhang, Qiang Ye, Wenying Zheng, Yijun Li

publication date:

December 2010

keywords:

Sentiment classification, Supervised approach, Semantic orientation approach, Chinese

abstract:

Sentiment classification aims at mining word-of-mouth, reviews of consumers, for a product or service by automatically classifying reviews as positive or negative. Few empirical studies have been conducted in comparing the different effects between machine learning and semantic orientation approaches on Chinese sentiment classification. This paper adopts three supervised learning approaches and a web-based semantic orientation approach, PMI-IR, to Chinese reviews. The results show that SVM outperforms naive bayes and N-gram model on various sizes of training examples, but does not obviously exceeds the semantic orientation approach when the number of training examples is smaller than 300.

© Atlantis Press. This article is distributed under the terms of the Creative Commons Attribution License, which permits non-commercial use, distribution and reproduction in any medium, provided the original work is properly cited.

full text:

0059_Ch057.pdf (150 K)

		back to author index

title:		Sentiment Classification for Consumer Word-of-Mouth in Chinese: Comparison between Supervised and Unsupervised Approaches
publication:		ICEBI-10
part of series:		Advances in Intelligent Systems Research
ISBN:		978-90-78677-40-6
ISSN:		1951-6851
DOI:		doi:10.2991/icebi.2010.56 (how to use a DOI)
author(s):		Ziqing Zhang, Qiang Ye, Wenying Zheng, Yijun Li
publication date:		December 2010
keywords:		Sentiment classification, Supervised approach, Semantic orientation approach, Chinese
abstract:		Sentiment classification aims at mining word-of-mouth, reviews of consumers, for a product or service by automatically classifying reviews as positive or negative. Few empirical studies have been conducted in comparing the different effects between machine learning and semantic orientation approaches on Chinese sentiment classification. This paper adopts three supervised learning approaches and a web-based semantic orientation approach, PMI-IR, to Chinese reviews. The results show that SVM outperforms naive bayes and N-gram model on various sizes of training examples, but does not obviously exceeds the semantic orientation approach when the number of training examples is smaller than 300.
copyright:		© Atlantis Press. This article is distributed under the terms of the Creative Commons Attribution License, which permits non-commercial use, distribution and reproduction in any medium, provided the original work is properly cited.
full text:		0059_Ch057.pdf (150 K)