Title:Clustering-based Partitioning for Large Web Graphs

Abstract:Graph partitioning plays a vital role in distributedlarge-scale web graph analytics, such as pagerank and labelpropagation. The quality and scalability of partitioning strategyhave a strong impact on such communication- and computation-intensive applications, since it drives the communication costand the workload balance among distributed computing this http URL, the streaming model shows promise in optimizing graphpartitioning. However, existing streaming partitioning strategieseither lack of adequate quality or fall short in scaling with alarge number of this http URL this work, we explore the property of web graph clusteringand propose a novel restreaming algorithm for vertex-cut parti-tioning. We investigate a series of techniques, which are pipelinedas three steps, streaming clustering, cluster partitioning, andpartition transformation. More, these techniques can be adaptedto a parallel mechanism for further acceleration of this http URL on real datasets and real systems show that ouralgorithm outperforms state-of-the-art vertex-cut partitioningmethods in large-scale web graph processing. Surprisingly, theruntime cost of our method can be an order of magnitude lowerthan that of one-pass streaming partitioning algorithms, whenthe number of partitions is large.