Hi! One really intriguing and fun example of what you can do with giraph is the algorithm of Backstrom and Kleinberg [1] that uses only network structure to order your friends according to their probability of being in a relationship with you. I have implemented it for Giraph (~50 lines of Java code) so I can share it if you think it is useful. One drawback for a presentation is that I am aware of only anonymized publicly available graphs, so the results would be something like node '2' is very likely to be in a relationship with node '23', etc. But you can always use results from [1] where they show that the accuracy of the algorithm is surprisingly high (over 0.7 if I remember correctly) and just focus on how easy and intuitive it is to write a parallel graph algorithm that can extract something interesting from a huge graph using commodity hardware.
In terms of performance, the giraph distribution features examples of algorithms such as PageRank or ShortestPaths. So perhaps you could show that you can easily handle a parallelizable task such as the PageRank algorithm for a huge graph. Executing PageRank was the reason that motivated us to use juju with giraph in [2], because the SNAP library runs in single-core and can't handle large graphs and Spark's GraphX requires 50x RAM which we didn't have... --Panagiotis [1] Lars Backstrom and Jon Kleinberg. 2014. Romantic partnerships and the dispersion of social ties: a network analysis of relationship status on facebook. In Proceedings of the 17th ACM conference on Computer supported cooperative work & social computing (CSCW '14). ACM, New York, NY, USA, 831-841. [2] Panagiotis Liakos, Katia Papakonstantinopoulou, Michael Sioutis, Konstantinos Tsakalozos, Alex Delis: Pinpointing Influence in Pinterest. SocInf@IJCAI 2016: 26-37 2017-03-07 15:52 GMT+02:00 Mark Shuttleworth <m...@ubuntu.com>: > Hi folks > > I have been invited to do a demo of a large-scale cloud deployment, so I > thought of our recent thread on Giraf. > > Would this be a great thing to show off, is it ready for wider > consumption? If we direct a bunch of people to try it out, will they > have a positive experience? Are there fun examples or benchmarks of > Giraf? Is there a good way to visualize what's going on? I've done a bit > of Hadoop in the past, but Giraf would be new for me. > > Mark > > > > -- > Juju mailing list > Juju@lists.ubuntu.com > Modify settings or unsubscribe at: > https://lists.ubuntu.com/mailman/listinfo/juju -- Juju mailing list Juju@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/juju