摘要： 互联网的发展使网页分类技术成为了研究热点，针对传统的基于统计的分类技术只能提供精确分类的情况，该文运用模糊综合评判理论进行文本分类，根据自然语言的复杂性和理解的不确定性，使分类后的网页以一定的概率分属于各个类别，真实地反映了网页信息。实践表明，用户可以根据所得概率进行灵活处理，以满足各种实际需要。

关键词: 网页分类, 特征提取, 模糊综合评判

Abstract: Web page categorization has become popular with the fast growing of Internet. Because of the disadvantages of traditional statistical approach, this paper proposes a categorization method based on fuzzy comprehensive evaluation. Due to the complexity of natural language and uncertainty of comprehension, the Web page belongs to every class with its probability, which can reflect the true information. According to the probability, users can deal with the Web pages flexibly to satisfy their needs.

Key words: Web page categorization, extraction of features, fuzzy comprehensive evaluation