PegasusGIM-V

GIM-V stands for Generalized Iterative Matrix-Vector multiplication.

Graph mining applications such as PageRank, spectral clustering, diameter estimation, connnected components etc. used repeated

matrix-vector multiplication

GIM-Vis highly optimized

It achieves (a) good scale-up on thenumber of available machines (b) linear running time on thenumber of edges, and

(c) more than5 timesfaster performanceover the non-optimized version ofGIM-V

web graphs can be downloaded from http://snap.stanford.edu/

Parse the graph file to get the input file format

Run that input file in Hadoop MapReduce jobs