Re: RDF Graphs

On 28 October 2014 03:50, Simon Spero <sesuncedu@gmail.com> wrote:

> I've never seen the real advantage of the graph metaphor for RDF; is it
> any easier to understand than viewing triples as a set of ground binary
> formula, or as a conjunction of binary formula embedded in a mess of
> existential qualifiers (because blank nodes)?
>
In my view there's 3 basic topologies, all of which lead to very different
types of user behaviour:

1. Tree like
2. Graph like
3. Network like

1. Tree like is what you see in most hierarchical systems.  The idea is
that you have a head node, and this inevitably leads to a degree of
centralization.  Most systems created by large companies are a tree, with
them near the top.  It's quite remarkable that the web actually took off
when tree structures were dominant.

2. Graph like, this is a tree without a head.  Much more peer to peer, tho
clusters do emerge and you get a degree of centralization, there is a long
tail.  If anyone has read weaving the web chapter 1, you'll see one of the
motivations of the web was that 'anything can be connected to anything'.
This is a topology that has changed the world.  We may be losing it tho.

3. Network like.  In mathematical graph theory a network is a graph where
the edges have numerical weights.  Sometimes described as pipes in a system
with different capacities of flow.  This leads to a whole new branch of
mathematics and computer science.  More importantly it offers a new dynamic
between nodes where information can be encouraged to flow through parts of
the network.  I would consider this a frontier more akin to an incentive
based graph structure.  Turns out this is pretty hard to model in RDF, so
we have less of it.

Just my 2 cents...


> Simon
>