Computer Engineering and Applications ›› 2018, Vol. 54 ›› Issue (14): 158-162.DOI: 10.3778/j.issn.1002-8331.1703-0332

Previous Articles     Next Articles

Text sentiment classification based on sentiment series distance and turning assimilation

ZHENG Cheng, TAN Xiaoyu, YU Xiukai, CAO Yang   

  1. School of Computer Science and Technology, Anhui University, Hefei 230601, China
  • Online:2018-07-15 Published:2018-08-06


郑  诚,谈小雨,于秀开,曹  杨   

  1. 安徽大学 计算机科学与技术学院,合肥 230601

Abstract: Considering the overall texts orientation identification of Chinese emotional texts and their sentiment express sequence have a great connection, and for the emotional texts, emotion tendency expression which is closer to the end of the texts is more influential to identify whole texts orientation. Therefore, for the short texts which express emotion inconspicuously or complicatedly, this paper considers the sequence of the appearance of the emotional nodes in the text and the sentiment assimilation of the turning point to classify the emotional texts more efficiently. The experimental result on the emotional texts dataset crawled from a shopping site shows that the classification method proposed in this paper is significantly better than the support vector machine classifier that is based on word features simply.

Key words: text sentiment classification, Support Vector Machine(SVM), sentiment distance, turning assimilation

摘要: 考虑到中文评价文本的整体情感倾向性与其表达的情感顺序有很大关系,且在具有情感倾向的中文文本中,越是靠近文本最后所表达的情感倾向,对于整个文本的情感分类影响越大。因此对于情感倾向表达不明显或者表达不单一的短文本,通过考虑文本中情感节点出现的顺序以及情感转折同化来对文本进行情感分类。在来自某购物网站爬取的中评评价文本数据集上的实验结果显示,提出的分类方法明显高于单纯基于词特征的支持向量机(SVM)分类器。

关键词: 文本情感分类, 支持向量机, 情感距离, 转折同化