Abstract
PageRank is a perfect solution that can rank pages in a mathematical based way. This paper introduces some important linear algebra concepts and graph theory, which are the basic of PageRank algorithm, elaborates the PageRank algorithm deeply based on these concepts, and show how PageRank algorithm ranking the web pages according to the hyperlinks pointing at them. Then it uses the method proposed by Taher H. Haveliwala, which focusing on improving the PageRank algorithm by using a set of PageRank vectors, which can prevent the problems of flawed topics in some degrees. However, this method may affect by worthless web pages, so this paper focuses on proposing a method that include two factors of time through users’ view point that could prevent spam web pages. At last, this paper conducts an experiment to show that this new model could add accuracy on searching results, and has practical utilities.