[p2p-research] the wikipedia decline

[J. Andrew Rogers]

On Thu, Nov 26, 2009 at 7:04 AM, Paul D. Fernhout
<pdfernhout at kurtz-fernhout.com> wrote:
> Our tax dollars at work. :-(
> http://www.pdfernhout.net/open-letter-to-grantmakers-and-donors-on-copyright-policy.html
> http://www.pdfernhout.net/on-funding-digital-public-works.html


The funding sources are about a quarter private, a quarter academic,
and half military/defense. The US military is way ahead of the curve
on developing real-world applications for semantic web technology, so
they are feeling most of the pain from existing weaknesses. Much of
what they currently use (and funded) is open source.


> But there is some open stuff out there. But they demand too much of users.


The closed stuff isn't any better. The difficulty reflects the state
of the technology and the workarounds/hacks currently required to give
us what we have.


> Related:
>  "Metacrap: Putting the torch to seven straw-men of the meta-utopia"


Generalized graph databases (semantic technologies are a type of graph
database) are an extremely powerful concept, but very limited in
implementation.  A single query on a database small enough to fit in
the memory of my laptop can take several minutes to run.

Doctorow is correct. Exhaustive use of metadata is a way to severely
restrict the scope of graph databases so that they scale better. Of
course, doing this makes them brittle and most of the value of
graph-like databases is that they can be schema-less. If you are going
to slap a fixed schema on them, you might as well use something else
that scales better...

A truly general semantic graph technology would change the way a lot
of things are done because you could analyze relationships that are
currently computationally intractable for all practical purposes. The
current set of algorithms central to graph-like databases neither
parallelize nor even scale well on single systems, putting most
real-world applications out of reach.


-- 
J. Andrew Rogers
realityminer.blogspot.com