Sun, XiaobaiQian, Yuchen2018-05-312018-05-312018https://hdl.handle.net/10161/17067<p>In modern network analysis, the PageRank algorithm has been used as an indispens-</p><p>able tool to determine the importance and relevance of the network nodes. Inspite</p><p>of extensive research conducted to accelerate the algorithm or its variants, there are</p><p>few studies about the effects of the damping factors on the ranking distribution.</p><p>To understand how the damping factor can affect the rank distribution in different</p><p>PageRank models, specifically, the directed surfer model by Brin and Page, and the</p><p>heat-kernel PageRank by Chung. We studied the ranking vector (steady state distri-</p><p>bution) under different damping factor values with each model. Enabled by efficient</p><p>batch calculation of the ranking vectors, we conducted systematic experiments to</p><p>measure the discrepancies of the distributions, explored and explained the capability</p><p>of adjusting the steady-state distribution via the change in damping factors. Experi-</p><p>mental results show that the steady-state distribution by Brin-Page model responses</p><p>non-linearly to the change in damping factor α, while by Chung’s heat-kernel model,</p><p>the damping factor β casts negligble effect on steady-state distribution. With this</p><p>phenomenon, Brin-Page model may be preferable over Chung’s model on utilizing</p><p>the non-linear relationship between the damping factor and steady-state distribution.</p><p>The relationship can be utilized also to find the propagation speed(damping factor)</p><p>from observations of two or more consecutive distributions.</p>Computer scienceDamping FactorDistributionKL divergencePageRankMaster's thesis