A novel content and style based measurement of web pages distance

Nowadays, many web-based systems have been using machine learning techniques in order to design more intelligent mechanisms for organizing, indexing, and retrieving web content, and it is necessary for more and more researches and applications to calculate the distance of web pages rationally. Generally proposed methodology is fit for extracting the differences between HTML documents of web pages, results of which cannot be used to tell the actual distance, between the content of web pages and the facade displayed in internet explorers. Based on these above, content distance, style distance, and hybrid distance are proposed in this paper, to make measurement result more practical. The efficiency will be proved through some classical experiments.