【作者】 高洁；

【导师】 吉根林；

【作者基本信息】 南京师范大学 ， 计算数学， 2004， 硕士

【摘要】 文本自动分类是指根据文本内容自动确定文本类别的过程。其目的是为信息检索提供更高效的搜索策略和更准确的查询结果。随着网络信息的快速增长，文本自动分类技术的研究对于网上信息搜索具有重要的意义。 本论文对中文文本自动分类技术作了系统论述，介绍了用于文本表示的向量空间模型、文本特征获取方法，较深入地讨论了基于支持向量机(SVM)的文本分类算法和Bayes文本分类算法。从提高训练速度，减少存储空间，充分利用历史信息的角度提出了增量式SVM文本分类算法。针对难以获得大量有类标签的训练集问题，提出了增量式Bayes文本分类算法。设计并实现了中文文本自动分类原型系统，对有关中文文本分类算法的有效性进行了验证分析。 文章详细分析了文档频率DF、信息增益IG、CHI统计和互信息MI的优缺点，提出了将文档频率DF和其它三种特征选择相结合的组合特征选择方法，实验结果表明组合的特征选择方法显著地提高了分类的精度。更多还原

【Abstract】 Text automatic categorization, the process of assigning one or multiple predefined category labels to free text documents, provides more effective search strategies and more exact query results for information retrieval. With the rapid growth of the information resources on Internet, it has become more and more important for text automatic categorization to search information on Internet.The thesis summarizes systematically techniques of Chinese text automatic categorization. Vector Space Model (VSM) which is used to represent text and feature acquiring methods are introduced. Categorization algorithms based on SVM and Bayes method are deeply investigated. A new incremental learning with SVM method is proposed to boost training speed, decrease storage space and use the history information fully. Incremental learning is an effective method for learning the classification knowledge from massive data, especially in the situation of high cost in getting labeled training examples, so an incremental Bayesian learning model is presented. An experimental system of Chinese text automatic categorization is built up to verify the validity of categorization algorithms those proposed above.Document frequency, information gain, mutual information and CHI statistic are analyzed in detail and compared through experiments. A combined feature selection method is proposed. The experimental results show that combined feature selection method can improve the classification precision.更多还原