为了有效地组织因特网上极其丰富的信息资源 ,通过分析中文和中文网页的特点 ,提出了一种新的中文网页的自动分类算法。这种算法主要利用字间的相关信息、词频以及页面的标记信息等 ,提取网页特征 ,并计算可调的词频加权参数 ,然后通过本类和非本类训练 ,建立专家数据库。实验表明 ,该算法可以获得 80 %以上的网页分类准确率 In order to effectively organize extremely rich information resources on the Internet, this paper proposes a new automatic classification algorithm for Chinese web pages by analyzing the characteristics of Chinese and Chinese web pages. This algorithm mainly uses the inter-word correlation information, word frequency and page marking information to extract the web page features and calculate the adjustable word frequency weighting parameters, and then build an expert database through this and non-class training. Experiments show that the algorithm can get more than 80% web page classification accuracy