Ulrik Brandes,
"A
faster algorithm for betweenness centrality."
This paper gives the efficient O(nv*ne) sequential BC algorithm
that you will parallelize for Homework 5. Spells out lots of
details.
Eric Robinson and Jeremy Kepner,
"Array
based betweenness centrality" (slides).
I used some of these slides in class on April 29. Slides 7-18 are
a very nice cartoon of one iteration of the outer loop of Brandes's
algorithm. The rest of the talk is about using sparse arrays and
sparse matrix primitives to parallelize BC.
HPC
Graph Analysis Benchmark.
The official specification for BC as a benchmark to measure TEPS =
traversed edges per second. Includes a C/OpenMP parallel code.