Survey On Graph Databases : Graph Database

3635 Words15 Pages
Survey on graph databases
Informatics, University of Edinburgh
Abstract. Graph databases, also called graph-oriented database, is a type of not only SQL (NoSQL) database based on graph theory that can store, map and query data relationships. Because this kind of database ensures its robust performance in processing graph-like data, it has been widely used in industry, for instance, Facebook and Twitter are using graph databases to store and analyze their user pro les. This paper re- viewed motivations and development of graph databases. Then it sur- veyed ve typical implementations (Apache Giraph, DEX/Sparksee, Hy- perGraphDB, Neo4j, Titan) of this kind of databases via four dimensions
(data model, storage system, query method, API and tools). The re- sults of comparison gure out the limits and strengths of di erent graph databases. 1 Introduction
In 1970 E.F.Codd published the paper A relational model of data for large shared data banks"[13] where the relational model was proposed. After decades of development Relational Database Management System (RDBMS) based on this model has become the most prevailing technology for the storage and anal- ysis of data today. However, because of the thrive new internet technologies such as web 2.0 sites[20], the growing number of internet users, as well as the increasing data sources like mobiles, sensors, the datasets developers face are becoming unprecedentedly large and complex that it is hard to process in tradi-
Get Access