A Common Subspace Construction Method in Cross-Domain Sentiment Classification

DOI

10.2991/esac-15.2015.13How to use a DOI?

Keywords

Cross-domain, Sentiment classification, Common subspace

Abstract

In this paper, we study the problem of domain adaptation in sentiment classification. Many existing approaches reduce the gap by extracting domain-independent topics. However these methods couldn’t cope with features which have different sentiments in different domains. To solve this problem, a common subspace construction method (CSC) is proposed in our paper. Firstly, the consistency of features' sentiment orientation in different domains is introduced to identify the common subspace. Then, domain-dependent features will be projected to this subspace. Empirical studies on benchmark tasks of sentiment analysis validate our assumption and demonstrated significant improvement of our method over competing ones in classification accuracies.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Yuhong Zhang
AU  - Xu Xu
AU  - Xuegang Hu
PY  - 2015/08
DA  - 2015/08
TI  - A Common Subspace Construction Method in Cross-Domain Sentiment Classification
BT  - Proceedings of the 2015 International Conference on Electronic Science and Automation Control
PB  - Atlantis Press
SP  - 48
EP  - 52
SN  - 2352-538X
UR  - https://doi.org/10.2991/esac-15.2015.13
DO  - 10.2991/esac-15.2015.13
ID  - Zhang2015/08
ER  -