Building more efficient database applications with a graph database

Graph Store

Article from ADMIN 08/2012

By Tim Schürmann

A graph database can identify the really important relationships on a social web, find the shortest paths, and optimize visitor flows. We compare some leading open source graph database options.

Peter knows Paula, a freeway links Cologne with Dortmund, and a network cable leads from Marvin to Zaphod. These relationships can be recorded in a visually intuitive way (Figure 1), and the results are referred to by computer scientists as graphs. In combination with the matching algorithms, graphs are perfect for computing routes, analyzing relationships (who knows whom?), identifying bottlenecks on networks, optimizing pipeline systems, or avoiding congestion on a freeway.

Figure 1: A simple graph shows the relationships between persons.

In the real world, application developers often crunch their graph data into relational databases, but using a traditional relational database can eat a lot of space and take a toll on performance, as well as burning lots of programming time. Graph databases are an infinitely preferable solution: they store the mesh of relationships, road maps, and networks in a space-saving way, thus supporting fast queries; some graph databases even offer efficient analysis algorithms.

Although graph databases have existed for many years, they have remained in the background until fairly recently. New developments such as NoSQL, the semantic web, and geodetic information systems such as OpenStreetMap (

...

Use Express-Checkout link below to read the full article (PDF).