本文阐述了一个基于K-Nearest Neighbor(KNN)算法的中文网页自动分类系统的体系结构、具体设计和实现过程。系统分成三个模块：中文网页的下载与过滤（王勋完成）；KNN算法的训练与分类（黄健完成）；算法评估以及分词系统的整合（吴亮完成）。

本人主要是实现算法的评估以及分词系统的整合：分词系统的整合利用爱博汉语分词系统(服务器版本—共享版)，将下载下来的中文网页过滤后所得到的txt文件内容进行分词，以便提供给KNN算法进行训练和分类；算法的评估是将KNN算法后所得到的结果进行分析，具体表现是通过查全率(r)、查准率(p)、宏观查全率、宏观查准率、微观查全率、微观查准率的值，以及对求出来的值以直方图、曲线图、梯形图直观的表现出来。

-This article is elaboration a system of Chinese page categorization based on K-Nearest Neighbor(KNN) algorithm while is architecture and designs as well as realizes the process specifically. The system are divides into three modules, Chinese homepage downloading and filtration (Wangxun completion), KNN algorithm training and classification (Huangjian completion), and algorithm evaluation and integration of Chinese word segmentation system (Wuliang completion).

I mainly is realizes what is algorithm evaluation and integration of Chinese word segmentation system. The Chinese word segmentation system is based on ACWPS that is will download after Chinese homepage filtration, and obtains the　.txt　document content carries on the participle. For it is providing for the KNN algorithm carries on the training and the classification. After the algorithm appraisal is the KNN algorithm, obtains the result carries on the analysis. Specific performance through the recall rate (r), poor-rate (p), m