Jordan Canonical Form of the Google Matrix: A Potential Contribution to the PageRank Computation
暂无分享,去创建一个
We consider the web hyperlink matrix used by Google for computing the PageRank whose form is given by A(c)=[cP +(1-c)E]T, where P is a row stochastic matrix, E is a row stochastic rank one matrix, and $c\in [0,1]$. We determine the analytic expression of the Jordan form of A(c) and, in particular, a rational formula for the PageRank in terms of c. The use of extrapolation procedures is very promising for the efficient computation of the PageRank when c is close or equal to 1.
[1] Carl D. Meyer,et al. Matrix Analysis and Applied Linear Algebra , 2000 .
[2] Gene H. Golub,et al. Extrapolation methods for accelerating PageRank computations , 2003, WWW '03.
[3] Taher H. Haveliwala,et al. The Condition Number of the PageRank Problem , 2003 .
[4] Gene H. Golub,et al. Exploiting the Block Structure of the Web for Computing , 2003 .
[5] Taher H. Haveliwala,et al. The Second Eigenvalue of the Google Matrix , 2003 .