World: r3wp

Now Ron Everett -- Associative Systems

associative data models

ah, interesting ...

perhaps it's like my Relations Engine, just done by someone who knows 
what he's talking about :-)

btw, sorry for the rushy session we are missing time.

its MUCH more   ;-)  beleive me  ;-)

(i recognise my spoken english is much to bad to follow this. hope 
there comes something written/recorded (if recorded i can try it 
a few times^^)

Max - I think henrik was talking about associative session :-)

His first slide said Relavance Technology.
That seems to be this:
http://www.associativesolutions.com/relavance.php

thanks..

why are we always getting back to xml when w have blocks and lists 
;-) funny...

yes, me to:-)

(pekr)

Because it's so nicely messed up.

sadly, xml is the csv of the web :-)

csv = complex structured values

Pekr: schedule has changed due to ...hem, you know, pb with Skype 
session and so on.

Carl said us at the beginning that he coul addadpt his speach to 
the needs. Making 2, 3 or more sessions, grouping things...

So modules where introduced this morning, but We will not have time 
to make a session on it.
Hope he will pubhished his slide on it anyway

Hey Ron, don't talk so much. Show us the system...

it comes

comma separated values: dumb data for interchange between systems

I like to see more of the sheets from Ron... goes pritty deep..nice...

sunanda,  specifically, that site is not the relavance site.  maybe, 
its a local reseller of the technology.

Except xml still can't pass "nulls" without SOAP or some such...sad

relavance is canadian technology.

sorry, max!

It's just using a graph-based approach. Good idea, and really better. 
no query necessary just traverse the graph.

wow they relay digged it all out... nice..

they actually have no website AFAIK.

www.e4graph.com

Adding database accessors at the leafs... well. Not so new up to 
now.

integrating with rebol though is a nice close fit based on linquistics..

robert you are not getting it... there is no tables or accessors.

Do they have graph-layout included?

it sounds very much like my RE, except my RE does not have that ability 
to move around freely in the structure, and you have to cross relate 
manually. no "traversing" back and forth. you can't start in the 
"middle" of the tree.

not sure, but they work at a much higher level, they can aggregate 
contexts and relationship set based on multi-dimensional.

henrik, nope, the concept of contexts is paramount.  you are working 
in one dimension, relavance has n dimensions.

pfiew...

Ron did his home work ...

What is that good for practically? Is there any commercial app using 
that aproach? Or is that only academic thingy? :-)

14 years of development.

what...!

great!

n Dimension = n entry / output edges

didn't you hear him ... he talked about that "small gang" at DARPA 
with

no, its n sets of edges.

input/output is one dimension.  its a relationship of dependency.

Well, than you have clustered your edges to a grouping node...

actually its "requirement"

free your mind of a graph... each node as multiple sets of edges.

and you can operate based on what those sets mean to each other.