Graph (Theory and Databases) is a nice overview of graph theory and databases by Pere Urbón-Bayes. I saw it first at: Alex Popescu’s site.
I do have a quibble about slide 14 with the usual graph showing progress towards Everything connected.
To your lower left are ontologies, RDF, Linked Data, Tagging, moving I suppose from less to more connected.
There is only one problem: Everything is already connected.
It doesn’t need electronic or other information systems for connections.
What is at issue is the representation of connections in electronic information systems.
The reason I emphasize that point is that all representations in electronic information systems are partial representations of some connections.
And as far as that goes, all representations do better with some aspects of connections than others.
For example, I don’t think that ontologies are further up the connection line than folksonomies.
Depends on your particular requirements as to which one is more connected
In fact, even unstructured text has connections between words. They are just implicit. Same applies to almost everything in our current info-sphere. In my opinion, this ongoing buzz around graphs is just a result of seeing and formalizing (or trying to formalize) these implicit connections using some graph notation language like RDF or Topic Maps.
I suppose Pere tries just to articulate (in his presentation) that the size of graph databases is exploding; number of nodes and edges the database should store are substantially bigger in future.
Yet another question is whether one agrees with Pere. One could easily think that the idea of linking information (and subjects) enables splitting one huge graph database into many smaller graph databases.
Comment by Aki — December 16, 2010 @ 3:30 am
Aki, Good point, I suspect the “one huge graph database” presumes a uniformity of semantics that does not exist.
Perhaps not. Simply because nodes are “connected” doesn’t mean the connection makes sense to everyone who sees it. Would not be a useful graph database but then I have seen other information resources that were not helpful. 🙂
Comment by Patrick Durusau — December 16, 2010 @ 5:36 pm