| 基于关系权重的文本表示法 |
| |
| 引用本文： | 倪洁琼,吴耿锋,郑宇.基于关系权重的文本表示法[J].计算机应用与软件,2009,26(5). |
| |
| 作者姓名： | 倪洁琼 吴耿锋 郑宇 |
| |
| 作者单位： | 上海大学计算机工程与科学学院,上海,200072 |
| |
| 基金项目： | 国家特种信息产品动员中心项目 |
| |
| 摘 要： | 提出一种基于关系权重的文本表示方法.通过优化关系权重,在文本向量中体现了不同特征项在不同类别中重要程度的差异,使得在此权重下不同类别的文本得到更准确的区分.运用SVM分类实验表明,基于关系权重的文本表示方法,较之传统的 TF-IDF 文本表示法,有更高的准确率和召回率.
|
| 关 键 词： | 自动文本分类 文本表示 关系权重 |
A RELATION-WEIGHT BASED DOCUMENT REPRESENTATION |
| |
| Ni Jieqiong,Wu Gengfeng,Zheng Yu.A RELATION-WEIGHT BASED DOCUMENT REPRESENTATION[J].Computer Applications and Software,2009,26(5). |
| |
| Authors: | Ni Jieqiong Wu Gengfeng Zheng Yu |
| |
| Affiliation: | School of Computer Engineering and Science;Shanghai University;Shanghai 200072;China |
| |
| Abstract: | We present a method of relation-weight based document representation.The method takes into account the text vectors that different feature items in different classes have different weightiness,so that the documents in different classes are classified better by optimizing the relation-weight.The experiments based on SVM classification show that our method has higher precision rate and recall rate in automatic text classification comparing with the method of TF-IDF document representation. |
| |
| Keywords: | Automatic text classification Document representation Relation-weight |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
