COLD. Revisiting Hub Labels on the Database for Large-Scale Graphs Full text

Alexandros Efentakis, Christodoulos Efstathiades, Dieter Pfoser
14th International Symposium on Spatial and Temporal Databases (SSTD 2015)
2015
Conference/Workshop
Abstract. Shortest-path computation is a well-studied problem in algorithmic theory. An aspect that has only recently attracted attention is the use of databases in combination with graph algorithms to compute distance queries on large graphs. To this end, we propose a novel, efficient, pure-SQL framework for answering exact distance queries on large-scale graphs, implemented entirely on an open-source database system. Our COLD framework (COmpressed Labels on the Database) may answer multiple distance queries (vertex-to-vertex, one-to-many, kNN, RkNN) not handled by previous methods, rendering it a complete solution for a variety of practical applications in large-scale graphs. Experimental results will show that COLD outperforms previous approaches (including popular graph databases) in terms of query time and efficiency, while requiring significantly less storage space than previous methods.