基于Hyperlink聚类的网页分类研究

Study of Web classification based on hyperlink clustering

研究并提出了基于Hyperlink聚类的分类算法 ,它不需要分析Web文档内容 ,只根据Web图来聚类 .算法性能比传统文本分类方法有很大提高 ,大大增强了网页分类的能力和效率 ,适合于海量网页分类 .实验表明基于Hyperlink聚类的分类算法 ,应用于Web文档信息分类 ,比传统的文本分类方法更加有效 .

This paper proposed a Web classification method based on Hyperlink clustering. It was not necessary to analyze the content and was more appropriate in huge Web page categorization than the traditional text clustering. It was showed that the application of classification method based on Hyperlink clustering to Web clustering was more effective and precise than the term-based clustering method.