Spectral clustering-based semi-supervised sentiment classification | |
Li, Suke ; Hao, Jinmei | |
2012 | |
英文摘要 | This work proposes a semi-supervised sentiment classification method. Our method utilizes spectral clustering-based algorithm to improve the sentiment classification accuracy. We adopt a spectral clustering algorithm to map sentiment units in consumer reviews into new features which are extended into the original feature space. One sentiment classifier is built on the features in the original training space, and the original training features combined with the extended features are used to train the other sentiment classifier. The two basic sentiment classifiers together form the final sentiment classifier through selecting instances in the unlabeled data set into the training data set. Experimental results show that our proposed method has better performance than Self-learning SVM-based sentiment classification method. ? Springer-Verlag 2012.; EI; 0 |
语种 | 英语 |
DOI标识 | 10.1007/978-3-642-35527-1_23 |
内容类型 | 其他 |
源URL | [http://ir.pku.edu.cn/handle/20.500.11897/325789] ![]() |
专题 | 软件与微电子学院 |
推荐引用方式 GB/T 7714 | Li, Suke,Hao, Jinmei. Spectral clustering-based semi-supervised sentiment classification. 2012-01-01. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论