PageRank

Overview

Pagerank algorithm implementation using Hadoop MapReduce and Java language.

Step

• According to the transition matrix input(transition.txt) build a relationship model
• Calculate the weight or transiton fact between pages
  • PageRank1 = Transition X PageRank0
• Sum up each unit weight to get new rank model
• Converge above steps N times

How to run

$ hadoop jar pagerank.Driver -trans /transition -rank /pagerank -unit /output -times 5

usage: pagerank.Driver

• -rank input rank file dir
• -times times of convergence
• -trans input transition file dir
• -unit unit output dir