DEX: A high-performance graph database management system

The amount of applications calling for efficient large graph management is dramatically increasing. Social network analysis, Internet or biocomputation are just three examples of such applications. In these cases, the interest focuses on the structural analysis of the relationships between different entities organized in huge networks or graph-like structures. Being able to efficiently handle such graphs becomes essential, placing graph database management systems in the eye of the storm. Among the different challenges posed by graph databases, finding an efficient way to represent and manipulate huge graphs that do not entirely fit in memory is still an unresolved problem. In this work, we present DEX, a high performance graph database management system based on bitmaps and other secondary structures. We show that by using bitmap structures for graph representation it is possible to improve the performance of a graph database system, allowing for the efficient manipulation of very large graphs containing thousands of millions of nodes and edges.