WOOster: a map-reduce based platform for graph mining
Aravindan Raghuveer · International Conference on Management of Data · 2011
Large scale graphs containing O(billion) of vertices are becoming increasingly common in various applications. With graphs of such proportion, efficient querying infrastructure becomes crucial. In this paper, we propose WOOster a hosted querying infrastructure designed specifically for the large graphs. We make two key contributions: a) Design of the WOOster framework. b)Scalable map-reduce algorithms for two popular graph queries: subgraph match and reachability. Our experiments show that the proposed map-reduce algorithms scale well with large synthetic datasets.