Leonid Zhukov Леонид Жуков

PageRank

Fast Parallel PageRank: Linear System approach to PageRank Computations

with David Gleich and Pavel Berkhin

Abstract:

In this paper we investigate the convergence of iterative stationary and Krylov subspace methods for the PageRank linear system and study the effect of the teleportation coefficient on its convergence. We demonstrate that linear system  converge faster than ''simple''  power iterations method and are less sensitive to the changes in teleportation. In order to perform the experiments we developed a framework for parallel PageRank computing. We describe the details of the parallel implementation and provide experimental results  obtained
on a 70-node Beowulf cluster.

Paper: Scalable Computing for Power Law Graphs: Experience with Parallel PageRank

Tex Report: Fast Parallel PageRank: A Linear System Approach

Power Point Presentation: Massively Parallel Scalable PageRank

big_picture2

blah

db-iteration-convergence db-time-convergence

blah

Picture6 convergence-c

Picture8

blah

table