Variable Damping Effect on Network Propagation
In modern network analysis, the PageRank algorithm has been used as an indispens-
able tool to determine the importance and relevance of the network nodes. Inspite
of extensive research conducted to accelerate the algorithm or its variants, there are
few studies about the effects of the damping factors on the ranking distribution.
To understand how the damping factor can affect the rank distribution in different
PageRank models, specifically, the directed surfer model by Brin and Page, and the
heat-kernel PageRank by Chung. We studied the ranking vector (steady state distri-
bution) under different damping factor values with each model. Enabled by efficient
batch calculation of the ranking vectors, we conducted systematic experiments to
measure the discrepancies of the distributions, explored and explained the capability
of adjusting the steady-state distribution via the change in damping factors. Experi-
mental results show that the steady-state distribution by Brin-Page model responses
non-linearly to the change in damping factor α, while by Chung’s heat-kernel model,
the damping factor β casts negligble effect on steady-state distribution. With this
phenomenon, Brin-Page model may be preferable over Chung’s model on utilizing
the non-linear relationship between the damping factor and steady-state distribution.
The relationship can be utilized also to find the propagation speed(damping factor)
from observations of two or more consecutive distributions.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Rights for Collection: Masters Theses
Works are deposited here by their authors, and represent their research and opinions, not that of Duke University. Some materials and descriptions may include offensive content. More info