This is an NP-hard problem 🙂 I can partially solve it and here is my up-to-date solution! Currently – and not for long – I am a master student at EMDC, doing my master thesis at Telefonica!
And same question goes again: What am I actually doing?
>> Bored to read? Just go directly to the bold line below – like a boss! <<
Well. We all know how big data is conquering the world. Big data analytics are important, rather to say necessary. So, mining large graphs – social graphs – can give us an insight of users’ behavior and preferences, market moves, and so on. Pregel is becoming popular for graph mining, thanks to its programming model – that makes it flexible – and its architecture – that makes it easy to scale. People have been putting efforts for building Apache Giraph – an open source implementation of Pregel. But, it is a “true story” the lack of open source implementation of algorithms on top of Giraph.
So! Why not writing some popular algorithms or even design new ones following the Pregel model? For each algorithm, I will follow a design-implementation-workload analysis-benchmarking process for having a holistic view of the algorithm’s behavior. I hope that I can contribute these algorithms to the open source community. Let’s see… 🙂
Long story short: I am creating an open-source library of graph mining algorithms on top of Apache Giraph.