Cover Image
QR Code

Parallel Graph Analytics.

The article explains data-centric abstractions and execution strategies to exploit parallelism in large-scale graph analytics. It describes the categories of graphs including planar graphs, social network graphs and random graphs and illustrates the concepts of data-centric abstraction of algorithms...

Full description

Published in:Communications of the ACM Vol. 59; no. 5; pp. 78 - 88
Main Authors: LENHARTH, ANDREW, NGUYEN, DONALD, PINGALI, KESHAV
Format: Article
Published: Association for Computing Machinery May2016
Subjects:
Planar graphs
Random graphs
Algorithms
Electronic data processing
Graph theory
Online Access:View this record in EBSCOhost
Summary:The article explains data-centric abstractions and execution strategies to exploit parallelism in large-scale graph analytics. It describes the categories of graphs including planar graphs, social network graphs and random graphs and illustrates the concepts of data-centric abstraction of algorithms using the Dijkstra and Bellman-Ford algorithms for the singe-source shortest path problem and collaborative filtering. It outlines vital choices in implementing parallel graph analytics programs.