Binary Information Press
With the rapidly growth of the World Wide Web, search engine has a lot of problems on how to quickly obtain the webpage which is the user most interested in. In order to find the user interested information from enormous web, the authors propose an effective approach - - A focused web crawling based on web semantic analysis and web link analysis. It is basis on the Formal Concept Analysis, using concept context graph calculate similarity between the web content and the users' interests. Then using the content similarity predict the similarity between the web links and the users' interests, this method only crawls the web pages which are related to users' interests, so the efficiency and precision of the crawls can be improved greatly.