基于bayes分类器的参考文献自动分类研究(附件)【字数:10536】
目录
摘要 3
关键词 3
Abstract 3
引言
一、 前言 4
(一)课题的背景及意义 4
1. 研究背景 4
2.应用前景 4
(二)国内外研究现状 4
1.国外研究现状 4
2. 国内研究现状 5
二、 相关研究 5
(一)向量空间模型 5
(二)经典朴素贝叶斯文本分类方法 6
1.贝叶斯定理 6
2.朴素贝叶斯分类 6
三、 实验步骤 7
四、 构造Bayes分类器 8
(一)建立数据集 8
(二)特征选择 9
(三)特征加权 10
(四)构造Bayes分类器 11
五、 实验结果评价 12
(一)评价方法 12
(二)实验评价 12
六、 结束语 13
(一)总结 13
(二)后续工作 13
致谢 15
参考文献 16
图 1 向量表达模型 6
图 2 朴素贝叶斯分类流程 7
图 3 文本预处理步骤 8
图 4 期刊论文原数据记录 9
图 5 部分有效数据集记录 9
图 *景先生毕设|www.jxszl.com +Q: ^351916072#
6 期刊论文特征选择后部分结果 9
图 7 学位论文特征选择后部分结果 10
图 8 会议论文特征选择后所得部分结果 10
图 9 软件类特征选择后所得部分结果 10
图 10 TFIDF计算后所得部分结果 11
图 11 bayes分类器研究结构 11
图 12 最终实验结果 13
表 1 测试文本与待定类之间的分类情况 12
表 2 实验结果 12
表 3 实验结果分析 12
基于Bayes分类器的参考文献自动分类研究
Automated classification of references
Student majoring in Information Management and Information System LuoMingni
Tutor Yang bo
Abstract:References are one of the important sources of statistical information on papers, so reference classification is a very important task. At present, a large number of submissions have led to problems such as lag in the classification of reference documents. In order to solve these problems, some scholars began to apply the automatic text classification technology to the classification of reference documents. After 1990, the automatic text classification method based on machine learning has gradually become the mainstream, and the naive Bayes classifier is a widely used text classifier, which uses statistical theory for text classification. The classification target of this research is mainly composed of the following techniques: text preprocessing, feature selection, feature weighting, etc., and introduces the methods of text classification and compares the advantages and disadvantages of each algorithm. Finally, the Bayesian classification method is selected for processing. . This paper analyzes and discusses the automatic classification of reference documents based on Bayes classifier, carries out a series of test work, and obtains experimental data. These experimental data show that this reference system can obtain good classification effect.
原文链接:http://www.jxszl.com/jsj/xxaq/564005.html