Link-Based Weighted PageRank

Abstract

PageRank has been known to be one of the most popular algorithms in Web ranking. In this algorithm, all links have been treated equally when distributing rank scores. This is not the case in the real world, since some Web resources are important than others and different links should be treated non-identically. In this paper, we adopt three strategies and propose the corresponding weighted algorithms to solve this problem. Numerical experiments illustrate the efficiency of our new algorithms.