“Graph databases are based on graph theory. Graph databases employ nodes, properties, and edges. Nodes are very similar in nature to the objects that object-oriented programmers will be familiar with.”
Wikipedia entry on Graph Databases.
I believe that Graph data models represent the likely future for OSS. And I’ve long held the belief that flexible object-oriented data structures are required for next-generation OSS.
When looking for material for this blog entry, I came across Project Rothko, which some of you may have already been familiar with. It was new to me but obviously aligns with the prior thinking discussed above.
As mentioned in previous entries, the complexity of integration is one of the major reasons why the OSS of large telcos are actually becoming bogged down – being more of a hindrance than a help. Project Rothko works on the principle of performing semantic search on data rather than performing structured data integrations (ie no schemas).
The data can remain in siloes but still relies on linking keys. From the linking data, learning algorithms can then be applied to monitor manual tasks and automate data connections. It can also potentially link data that is rarely integrated (eg service impact analysis with service levels with inventory audits with billing, etc)