A Semi-clustering Scheme for Large-Scale Graph Analysis on Hadoop

With the evolution of IT technologies, large-scale graph data have lately become a growing interest. As a result, there are a lot of research results in large-scale graph analysis on Hadoop. The graph analysis based on Hadoop provides parallel programming models with data partitioning and contains iterative phases of MapReduce jobs. Therefore, the effectiveness of data partitioning depends on how the data partitioning maintains data locality in each node of cluster. In this paper, we propose a semi-clustering scheme for large-scale graph analysis such as PageRank algorithm on Hadoop and show that the proposed scheme is effective. With experiment results, PageRank computation with the semi-clustering improves the performance.