电子学报 ›› 2017, Vol. 45 ›› Issue (4): 1018-1024.DOI: 10.3969/j.issn.0372-2112.2017.04.034

• 科研通信 • 上一篇    下一篇

基于情感标签的极性分类

周孟, 朱福喜   

  1. 武汉大学计算机学院, 湖北武汉 430072
  • 收稿日期:2015-10-26 修回日期:2016-03-07 出版日期:2017-04-25
    • 通讯作者:
    • 朱福喜
    • 作者简介:
    • 周孟 男,1986年生于河南濮阳,博士生、CCF学生会员,主要从事数据挖掘及自然语言处理方面的研究工作.E-mail:angel19851229@163.com
    • 基金资助:
    • 国家自然科学基金 (No.61272277)

Polarity Classification Based on Sentiment Tags

ZHOU Meng, ZHU Fu-xi   

  1. Computer School, Wuhan University, Wuhan, Hubei 430072, China
  • Received:2015-10-26 Revised:2016-03-07 Online:2017-04-25 Published:2017-04-25
    • Supported by:
    • National Natural Science Foundation of China (No.61272277)

摘要:

情感极性分析是文本挖掘中一种非常重要的技术.然而在不同领域中,很多情感极性分类系统存在分类精度低和缺少大量标注数据的缺陷.针对这些问题,提出了一种基于情感标签的极性分类方法.首先通过所有文本建立Sentiment-Topic模型,抽取出文本的情感标签;然后利用情感标签将文本划分为两个子文本,并通过Co-training算法对子文本进行分类;最后合并两个子文本的分类结果,并确定文本的情感极性.实验结果表明该方法具有较高的分类精度,而且不需要大量的分类样本.

关键词: 极性分类, 情感标签, 半监督学习, co-training学习

Abstract:

Sentiment analysis is a very important technology in text mining.However,a number of systems require amounts of annotated training data in different fields.In order to solve these problems,an approach to polarity classification based on sentiment tags is proposed.Firstly,on the basis of all the documents,the sentiment-topic model is developed and the sentiment tags for each review are extracted.Then each review is divided into two sub-texts by these sentiment tags,and each sub-text is classified by exploiting the co-training algorithm.Finally,the category results of two sub-texts are combined to determine document-level polarity of each review.Experimental results show that compared with other algorithms,the method improves the classification precision without a large number of annotated samples.

Key words: polarity classification, sentiment tag, semi-supervised learning, co-training learning

中图分类号: