Dynamic Partitioning of Large Scale RDF Graph in Dynamic Environments

In this paper, we propose a dynamic partitioning scheme of RDF graph to support load balancing in the dynamic environment. It generates clusters by grouping RDF subgraphs with high use frequencies while generating subclusters with RDF subgraphs with lower use frequencies. These clusters and subclusters perform load balancing based on the mean frequency of queries for the distributed server. In addition, the number of edge-cuts connected to clusters and subclusters is minimized to minimize the cost of communication between servers.