摘要： 文本分类技术是文本数据挖掘的基础和核心，是基于自然语言处理技术和机器学习算法的一个具体应用。特征选择和分类算法是文本分类中两个最关键的技术，该文提出了利用潜在语义索引进行特征提取和降维，并结合支持向量机(SVM)算法进行多类分类，实验结果显示与向量空间模型(VSM)结合SVM方法和LSI结合K近邻(KNN)方法相比，取得了更好的效果，在文本类别数较少、类别划分比较清晰的情况下可以达到实用效果。

关键词: 特征提取, 潜在语义索引, 支持向量机

Abstract: Text classification is the foundation and crucial problem of text data mining, it is an application based on the technology of natural language processing and machine learning. Feature extraction and categorization algorithm are the most crucial technologies for this problem. This paper proposes that latent semantic indexing (LSI) is used for feature extraction and dimensionality reduction, support vector machine(SVM) is used for text classification. The result shows that compared with the classifier based on vector space model combined SVM and the classifier based on LSI combined K-nearest neighbor (KNN), better performance is acheived. It shows that while the number of categories is small, and the categories are divided distinctly, the method can be used for practical application.

Key words: feature extraction, latent semantic index(LSI), support vector machine(SVM)

中图分类号: 

• TP391