Some more brainstorming following our discussion on Thuesday

Hello Ashok, everyone,

Regarding our last conversation I have been doing some research on property graphs and RDF.
Actually, an interesting fact is that in RDF 1.1 blank nodes are removed http://www.w3.org/TR/rdf11-concepts/#section-skolemization in order to create more native graphs.
This is of course just a nice information that would be helpful in the RDF serialisation, but not in the core of this CG.

The whole idea for using such a model and standardise it, is for creating applications (like google did with gmail and json-ld, https://developers.google.com/gmail/actions/getting-started) and make it easier for developers to create services with some interoperable graph schema.

We could even inspire ourselves with what is happening in the Linked Data community (where json-ld and RDF 1.1 are core part of) and find similar use cases and scenarios, of why would someone require such a graph schema?
For example, if neo4j (which is one of the most popular property graph databases) or other similar databases supported such a schema as an API, then it would making it really simple to create apps that query multiple such resources  and create new services on top of those.
It could even enable a new ecosystem of apps, like json-ld when adopted by google and other companies.

A really simple query that comes in mind, is for example:
Bring me this subgraph from Facebook that i want to correlate with this subgraph with my logged in user from instagram.

We could enable such capabilities.

This mail may not add  anything new in the discussion, but I am trying to summarise our work until now and find a way to improve our scope.

Regards,
Michael Petychakis.