| 基于改进的TF-IDF与隐朴素贝叶斯的情感分类研究 |
| |
| 引用本文： | 李晓东,肖基毅,邹银凤. 基于改进的TF-IDF与隐朴素贝叶斯的情感分类研究[J]. 南华大学学报(自然科学版), 2019, 33(2): 79-84 |
| |
| 作者姓名： | 李晓东 肖基毅 邹银凤 |
| |
| 作者单位： | 南华大学 计算机学院,湖南 衡阳 421001,南华大学 计算机学院,湖南 衡阳 421001,惠州工程职业学院,广东 惠州 516023 |
| |
| 基金项目： | 南华大学研究生科学基金项目(2018KYY087) |
| |
| 摘 要： | 为了提高情感分类准确率,提出了一种基于改进的TF-IDF与隐朴素贝叶斯的情感分类研究。通过改进的TF-IDF算法提取文本特征词,并根据属性之间的依赖关系添加隐藏的父节点,增强了属性之间的依赖关系,提高了情感分类的准确性。实验结果表明,在平均宏查准率、宏查全率和宏F1值在改进之后的算法分别提高了5%、8%和6%。
|
| 关 键 词： | 情感分类 隐朴素贝叶斯 TF-IDF 权重 朴素贝叶斯 |
| 收稿时间： | 2018-10-08 |
Research on Emotion Classification Based on Improved TF-IDF and Hidden Naive Bayes |
| |
| LI Xiaodong,XIAO Jiyi and ZOU Yinfeng. Research on Emotion Classification Based on Improved TF-IDF and Hidden Naive Bayes[J]. Journal of Nanhua University(Science and Technology), 2019, 33(2): 79-84 |
| |
| Authors: | LI Xiaodong XIAO Jiyi ZOU Yinfeng |
| |
| Affiliation: | School of Computer,University of South China,Hengyang,Hunan 421001,China,School of Computer,University of South China,Hengyang,Hunan 421001,China and Huizhou Engineering Vocational College,Huizhou,Guangdong 516023,China |
| |
| Abstract: | In order to improve the accuracy of emotion classification,it proposes an improved TF-IDF and hidden naive Bayes based emotion classification research.The improved TF-IDF algorithm is used to extract the text feature words and add hidden parent nodes according to the dependency relationship between attributes,which enhances the dependency relationship between attributes and improves the accuracy of emotion classification.The experimental results show that the improved algorithm increases the average macro precision,macro recall and macro F1 by 5%,8% and 6%,respectively. |
| |
| Keywords: | emotion classification hidden naive Bayes TF-IDF weight naive Bayes |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《南华大学学报(自然科学版)》浏览原始摘要信息 |
|
点击此处可从《南华大学学报(自然科学版)》下载全文 |
|
