PageSim

Abstract

To find similar web pages to a query page on the Web, this paper introduces a novel link-based similarity measure, called PageSim. Contrast to SimRank, a recursive refinement of cocitation, PageSim can measure similarity between any two web pages, whereas SimRank cannot in some cases. We give some intuitions to the PageSim model, and outline the model with mathematical definitions. Finally, we give an example to illustrate its effectiveness.