Question

毕业论文的摘要求个翻译。不要GOOGLE自动翻译的，我毕竟过了4级我能看出来。附上几个专业词汇。100分！

文本分类：Text categorization
文档频率(DocumentFrequency，DF)、互信息(Mutual Information，MI)、统计法(CHI—square，CHI)和信息增益（Information Gain :IG））
特征降维 Feature reduction

特征提取 Feature selection
文本分类中特征选择方法的比较

摘要

现在的我们处于一个科技高速发展的时代，一个信息爆炸的时代，每天有海量的信息出现在我们生活的各个角落，如何有效高速的处理这些信息，便成为了当下的研究热点。由此，文本自动分类应运而生。
特征选择是文本分类中一个十分重要的步骤。从原始文本特征集合中选择最佳的特征子集，是特征选择的目标。而对文本进行降维处理可以在保证分类准确性的前提下最大程度的减少开销、提高运行速度。降维的主要方法有两种，分别是特征选择和特征提取。本文介绍了在实际运用中常用的四种特征选择方法，即文档频率档频率(DF)、互信息(MI)、统计法(CHI)和信息增益（IG），并在中文文本语料库的基础上，分别对这四种方法进行了实验，从而得出了比较结论。实验结果表明，在综合效果考虑方面，文档频率的效果最佳。

关键字：文本分类；特征选择；文本降维；文档频率

Answer 1

Now we're in a high-speed development of science and technology era, an era of information explosion, a mass of information appears in every corner of our lives every day, how to deal with these information effectively and rapidly, it has become a research hotspot at present. Thus, automatic text classification emerge as the times require. Feature selection is an important step in text categorization. Selection of optimal feature subset from the original feature set, is the goal of feature selection. But the text dimension can guarantee to reduce expenses, increase the running speed of the maximum classification accuracy under the premise of. There are two main methods for dimensionality reduction, feature selection and feature extraction are. This paper introduces the actual application in four kinds of commonly used feature selection methods, namely the document frequency shift frequency (DF), mutual information (MI), statistics (CHI) and information gain (IG), and based on Chinese text corpus, respectively, the four methods are obtained from experiments, and the comparison of results. The experimental results show that, considering in synthetic effect, document frequency is best.