计算机应用与软件2016，Vol.33Issue(9)：301-305,5.DOI:10.3969/j.issn.1000-386x.2016.09.070
基于多约简 Fisher-VSM 和 SVM 的文本情感分类
TEXT SENTIMENT CLASSIFICATION BASED ON MULTI-REDUCED FISHER-VSM AND SVM
邢玉娟 1谭萍 1曹晓丽1
作者信息
- 1. 兰州文理学院数字媒体学院 甘肃 兰州 730000
- 折叠
摘要
Abstract
We propose a novel text sentiment classification algorithm in this paper,it is based on multi-reduced Fisher-VSM and SVM,to improve the accuracy of text sentiment classification.The algorithm first adopts Fisher’s discriminant criterion to extract TF-IDF eigenvector, and then clusters the documents according to the similarity between vector space models of low-dimension documents so as to reduce their numbers.The algorithm makes reduction on vector space model of documents from two aspects of dimensionality and number so as to improve the training speed and classification performance of SVM.Simulation experimental results demonstrate that the proposed algorithm has good re-call ratio and classification accuracy.关键词
文本情感分类/Fisher 判别比/向量空间模型/支持向量机Key words
Text sentiment classification/Fisher discriminant ratio/Vector space model (VSM)/Support vector machine (SVM)分类
信息技术与安全科学引用本文复制引用
邢玉娟,谭萍,曹晓丽..基于多约简 Fisher-VSM 和 SVM 的文本情感分类[J].计算机应用与软件,2016,33(9):301-305,5.
