基于k最近邻居算法文本分类器设计论文.doc

基于k-最近邻居算法文本分类器设计论文 基于k-最近邻居算法的文本分类器设计 摘 要：K－最近邻居算法是指存在一份要求分类的新样本，可通过在存在的训练样本中找出K个 与其最相近的样本加以分析，得出待分类样本的类别就是这K个样本中占比例最大的类别。K－ nearest文本分类器是根据K－最近邻居算法设计的以解决当今越来越急需的文本分类的工具，本文 在此设计过程中严格按照软件工程所要求的步骤来分析和实现及测试，并对测试项目作详细说明。 关键词：数据挖掘；分类算法；K-最近邻居算法；文本分类 Abstract: K-nearest algorithm is that finding the similar stylebook which request classifying in training stylebook, the most proportion of sorts in these stylebook is the sort which waiting for classifying stylebook. K-nearest sorter can solve now the more and more urgently needed text classification. K-nearest sorter designing process analyzes and realize that strictly according to the step which the software engineering requests, meanwhile has made the test to this sorter, and to test the item to give the explanation. Keywords: Data mining; classification algorithm; K-nearest algorithm; Text categorization 目 录 前 言 ..................................................................... 2 ......................................................... 4 1.1 文本分类简介 .................................................................................................................................... 4 1.2 文本的表示 ........................................................................................................................................ 4 1.3 文本分类的过程 ................................................................................................................................ 5 1.4 K－最近邻居算法 .............................................................................................................................. 6 1.4.1 算法介绍 ................................................................................................................................... 6 1.4.2 在文本中的应用 ....................................................................................................................... 7 ........................................ 8 2.1 K-NEAREST分类器的需求分析 ......................................................