| 基于朴素贝叶斯的中文海事文本多分类器研究 |
| |
| 引用本文： | 袁文生,王晓峰. 基于朴素贝叶斯的中文海事文本多分类器研究[J]. 计算机与现代化, 2011, 0(5): 150-153. DOI: 10.3969/j.issn.1006-2475.2011.05.043 |
| |
| 作者姓名： | 袁文生 王晓峰 |
| |
| 作者单位： | 上海海事大学信息工程学院,上海,200135 |
| |
| 摘 要： | 设计一个有效地基于朴素贝叶斯的中文海事文本多分类器。在文本分类的预处理步骤中,在中文分词上选取领域词典和停用词典有效地降低特征维数、选取IG特征提取方法、改进的TF-IDF公式中特征词权重的计算方法,以建立词频矩阵等,最后用选取的海事样本数据进行训练建立分类库。实验数据表明,本文的基于朴素贝叶斯的中文海事文本多分类器具有很好的高效性和准确性。
|
| 关 键 词： | 海事 文本分类 朴素贝叶斯 多分类 |
Research on Chinese Maritime Muti-class Text Classifier Based on Nave Bayes |
| |
| YUAN Wen-sheng,WANG Xiao-feng. Research on Chinese Maritime Muti-class Text Classifier Based on Nave Bayes[J]. Computer and Modernization, 2011, 0(5): 150-153. DOI: 10.3969/j.issn.1006-2475.2011.05.043 |
| |
| Authors: | YUAN Wen-sheng WANG Xiao-feng |
| |
| Affiliation: | YUAN Wen-sheng,WANG Xiao-feng(Information Engineering College,Shanghai Maritime University,Shanhai 200135,China) |
| |
| Abstract: | This paper designs a effective multi-class text classifier based on Nave Bayes for maritime area.In preprocessing of text classification,the feature dimensions are greatly decreased by combining professional dictionary and stop-words after Chinese Word Segmentation,this paper uses IG feature selection,improved TF-IDF to calculate feature words weights.Finally,the classification database for training Bayes is built.The experiment shows this maritime multi-class text classifier has great efficiency and accura... |
| |
| Keywords: | maritime text classification Nave Bayes multi-class text classification |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
