For a long time, Neo4j occupied a unique niche in the market with its graph based data model, but as data flows have become more complex, more and more vendors have been experimenting with the power of the graph. This week, open source search provider Elastic has stepped up with new capabilities for two of its offerings. Labelled simply as ‘Graph‘, it unlocks new potential for relationship analysis.
Along with Elasticsearch, Elastic has also added Graph to data visualisation plug-in Kibana, which allows developers to integrate their data into custom dashboards for easy insight sharing. Unfortunately, for now, it’s only available as part of Elastic’s existing subscription model, but if you’re eligible, you can also play with it on a trial basis.
Graph databases have come into their own as user demands for big data have progressed beyond merely streaming vast tracts of information to actually being able to unlock significant value from massive flows of information. The graph model facilitates new ways of looking at data and charting relationships that may otherwise be obscured. With their offering working in tandem with Elasticsearch, the company states that users can can perform a number of useful tasks. This includes things like detecting sources of risk, ie. analysing any shared behaviours of potential website hackers or real-time content recommendation (for example, bringing up a host of other potential items of interest if you brought something like a rubber unicorn mask). It can also be used to identify relationships between users, eg. how many Scala developers on Stack Overflow also use Kotlin.
While traditional search relationship mapping is grounded on calculating the frequency in which a relationship appears, Graph in Elasticsearch prioritises the significance of relationships versus the global average, generating more powerful results.
Although the tool clearly steps on the toes of others before it, Steve Kearns, senior director of product management for Elastic, is quick to note that they aren’t looking to conquer new market territory. Speaking to Datanami, Kearns commented, “I don’t expect this to put Hadoop out of business or to put any graph database out of business…I do expect people to be solving a new class of problem with a much smaller set of technology. Whereas before you might have had to have a Hadoop cluster to answer a question like a recommendation engine, now you don’t need that. You may decide you need to go there [Hadoop] for certain reasons, but it won’t be because you can’t do it on Elasticsearch, which in many cases is where this data already lives.”