An introduction to graph databases
Chloe Green talks about some of the graph databases that have been developed today, including Neo4j
One of the few companies selling a graph database platform commercially is Neo Technology.
Back in 2000, the company’s co-founders were building a new interface for a content management system (CMS). They wanted to use the connections between content stored in the CMS to help users find the information they needed, but they found that the relational database they were using did not make this at all easy.
“They realised that the [CMS] software was managing not just a lot of individual, isolated and discrete data items, but the connections between them,” explains Jim Webber, Neo Technology’s chief scientist.
“That connected data was more challenging to store in a relational table and tremendously slow to query.
“They sensed that, armed with a graph data model, our development team might not waste half its time fighting the database,” Webber says. “A few years later they found they could remove the underlying database, and created the world’s first modern graph database.”
An open source community version of that database, called Neo4J, was released in 2007, under an open source licence. Neo Technology also sells a commercial version, providing support and services to its paying customers.
Graph databases may be most commonly associated with social networks, but Webber says companies used Neo4J to analyse patterns in datasets ranging from telecommunications networks to genes and proteins.