Here we try to make a list of all systems targeting big graph analytics. All systems appear in some academic papers at some point. Open source implementations are preferred.
We annotate each system with links, paper and some highlights.
Please let us know if something is missing. We appreciate it a lot.
We view analytical platforms as in two categories, programming platforms and graph databases. The line between the two is bluring. Programming platforms interact with user via programming languages (scripts), therefore are much more flexible. Graph databases achieve that via graph query languages, in a more controled, concise way. Declarative languages may play a bigger role in the field later, but only with the functionalities (such as aggregation) added; just like what happend in OLAP systems.
Major players in the field are MapReduce model and vertex-centric programming model.
Though many people argue that MR is not the suitable platform for graph algorithms, MR is still the most common parallel infrastructure people have access to. The ability to handle billion node graph in a reasonable amount of time is good enough in many cases. Insights that are gained from designing in MR algorithms are certainly useful for other platforms (and vice versa).
Definitely the biggest player in the field.
White, Tom. Hadoop: the definitive guide. O'Reilly, 2012.
A full stack platform, from UC Berkeley. Also has a vertex-centric programming model.
Zaharia, Matei, et al. "Spark: cluster computing with working sets." Proceedings of the 2nd USENIX conference on Hot topics in cloud computing. 2010.