Graph database partitioning: A study
Ali Ben Ammar · 2016
Today, graph databases (GDB) represent a requirement for many applications that manage graph-like data, such as social networks. They are able to manage highly interconnected data such as analyzing whole-graph and answering user queries, which are more interested in the relationships between data rather than on the nodes of the graph. Partitioning data over several systems is one of the most techniques used to optimize queries. It allows distributing data over several servers when it is infeasible to query and store them on a single site. The aim is to improve query response time and to minimize the data storage cost. Although, it has been extensively used in traditional databases, data partitioning has many specificities when it is applied in GDB, which are characterized by a dynamic structure and highly interconnected data. In this paper, we study the recent approaches of GDB partitioning. We summarize and then discuss what partitioning options these approaches have used, what partitioning criteria they have optimized and what are the main steps for partitioning GDB.