基于文本的分类算法研究.docVIP

摘 要 随着互连网的飞速发展和信息传播手段的不断进步，各种文本数据已经源源不断地大量涌现，如文章摘要、电子邮件、网上即时消息等。在这些庞大的信息资源中，蕴含着大量的有价值的知识。为了能够从海量、异构和动态的文本资源中快速有效地发现所需的知识。在这样的背景之下，文本的自动分类技术正逐渐成为一个日益重要的研究领域。文本分类就是将自然语言表示的文本，根据其内容，自动分类到预先设定的类型集合中，使得一篇文本对应类型集合中的一个或者多个类型。 本文分析了文本分类相关技术，包括文本分类的一般过程、向量空间模型、特征提取、分类算法评价指标等。本文介绍了常见的文本分类算法，重点研究了KNN和贝叶斯分类算法及相关改进算法，并分析了算法的效果。 关键词：文本分类 SVM模型 KNN算法 Abstract In this paper several methods of text clustering was simple, the comparison of their advantages and disadvantages, in this based on the analysis of the impact factor of several methods, in particular, the k-mean method introduced improvements. With the rapid development of the Internet and information dissemination means progress, various essay this data has been flowing springing, such as writing the, E-mail, online instant messaging, etc. In order to apply them, there have been many statistical classification and machine learning method has been applied to the text classification, and have made quite a good effect, including: vector space model, recently K neighbor method, decision-making tree model, simple bayesian, support vector machine and neural networks. These methods a high degree of automation, stable performance, strong adaptability and relative to artificially text classification more efficient. This paper on the basis of text classification, made a in-depth research, focusing on this data short put forward effective classification algorithm. This paper analyzes the text categorization related technologies, including general process, the text categorization vector space model, feature extraction, classification algorithm evaluation index, etc. This paper introduces the common text classification algorithm, focus on the KNN and bayes classification algorithm and associated the improved algorithm, and the effect of the algorithm is analyzed. Keywords: text categorization SVM model KNN algorithm 目 录 摘 要 IV Abstract V 目 录 I 第一章 绪 论 3 1.1 课题研究的背景 3 1.2 文本分类的研究意义 4 第