| 改进的基尼指数在文本分类中的应用研究 |
| |
| 引用本文： | 唐伟,刘丰年,陈崇帮,欧新良,王苏. 改进的基尼指数在文本分类中的应用研究[J]. 长沙大学学报, 2013, 0(5): 55-57,63 |
| |
| 作者姓名： | 唐伟 刘丰年 陈崇帮 欧新良 王苏 |
| |
| 作者单位： | [1]湖南工业大学计算机与通信学院,湖南株洲412007 [2]长沙大学计算机科学与技术系,湖南长沙410022 |
| |
| 基金项目： | 湖南省自然科学基金（批准号：11JJ3002）资助项目;湖南省教育厅科技重点项目（批准号：09A010）. |
| |
| 摘 要： | 随着网上信息的极大丰富,文本分类技术显得越发重要,且预处理技术已成为文本分类的瓶颈.在预处理中采用TF-IDF算法,并且根据基尼指数的纯度原理对传统的基尼指数方法进行了基尼指数测度函数的改进,以降低原始文本的特征选择空间的维数.通过对比实验数据,表明这种改进是可行且有效的,体现在时间、空间复杂度小,精确度高.
|
| 关 键 词： | 文本分类 TF-IDF算法 基尼指数 测度函数 纯度原理 |
Application of Improved Gini Index in the Text Classification |
| |
| TANG Wei,LIU Fengnian,CHEN ChongbangI,OU Xinliang,WANG Su. Application of Improved Gini Index in the Text Classification[J]. Journal of Changsha University, 2013, 0(5): 55-57,63 |
| |
| Authors: | TANG Wei LIU Fengnian CHEN ChongbangI OU Xinliang WANG Su |
| |
| Affiliation: | 1. College of Computer and Communication, Hunan University of Technology, Zhuzhou Hunan 412007, China; 2. Department of Computer Science and Technology, Changsha University, Changsha Hunan 410022, China) |
| |
| Abstract: | In this paper, TF - IDF algorithm is used in text preprocessing, and the Gini coefficient measure function of the traditional Gini coefficient method is improved according to the purity principle of Gini coefficient so as to reduce dimensions of the original text feature space. Through comparing the experimental data, it is indicated that the improvement is feasible and effective, which is reflec- ted by the facts that the complexity of time and space is small and tht~ nreei~ir, n i~ h;~h |
| |
| Keywords: | text categorization TF - IDF algorithm Gini coefficient measure function the purity principle |
| 本文献已被 维普 等数据库收录！ |
|
