«上一篇/Previous Article|本期目录/Table of Contents|下一篇/Next Article»

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

期数:

2013年01期

页码:

112-114

栏目:

智能、算法、系统工程

出版日期:

1900-01-01

文章信息/Info

Title:

Research of Text Sentiment Classification Based on Kernel Fisher Discriminant

文章编号:

1673-629X（2013）01-0112-03

作者:

邢玉娟; 李恒杰; 曹晓丽; 张成文

甘肃联合大学 电子信息工程学院

Author(s):

XING Yu-juan; LI Heng-jie; CAO Xiao-li; ZHANG Cheng-wen

关键词:

文本情感分类; 核; Fisher; 判别; 支持向量机; 向量空间模型; Fisher; 线性判别

Keywords:

text sentiment classification; kernel Fisher discriminant; support vector machine; vector space model; Fisher linear discriminant

文献标志码:

A

摘要:

针对文本情感分类准确率不高的问题,提出基于核 Fisher 判别的文本情感分类方法,判别文本观点是正面还是负面.首先采用向量空间模型对文档进行数据化表示,然后将不同权重计算方法和词性特征选择规则与核 Fisher 判别方法相结合来判别文档的情感观点.实验结果表明:核 Fisher 判别方法在训练的过程中使用了所有的文本特征向量而不是少数几个支持向量,因此比传统支持向量机具有较高的分类准确率,同时不同的权重特征计算方法和词性特征的选取规则对文本情感分类准确率具有较大的影响

Abstract:

In view of the problem of low accuracy rate of text sentiment classification,the method of text sentiment classification based on kernel Fisher discriminant was proposed to decide that the text view is positive or negative. VSM was used to digitalize text firstly,and then classification results was achieved based on KFD combined with different weight computing method and different parts of speech fea-ture selection rule. The experiment results showed that KFD was superior to SVM in text sentiment classification problem, and also showed that the method of weight computation and the rule of parts of speech feature selection had big affection on recognition results

常用功能

更新日期/Last Update: 1900-01-01