There is now first cut access to TheBrain's .brn files from Nooron
through PyOKBC. As expected, .brn files have so little structure that
teasing out the semantics of a random brain file relies on various
fallible heuristics. The point was to be able to publish the contents
of .brn files using Nooron and that objective is now basically
achieved. Areas for improvement include:
1) initial caps before smashing words together for auto WikiWord frame
names
2) remove more junk characters to make better WikiWords
3) actually create slot frames
4) smash together words but with an initial lowercase for slotnames
(ala OWL)
5) make links between nodes
6) reify some slot values (e.g. Languages in the world_encyclopedia)
7) finally implement on-the-fly file format conversion (01)
These sample brains:
http://www.nooron.org/know/latin_to_english_phrases
http://www.nooron.org/know/world_encyclopedia
http://www.nooron.org/know/elements
http://www.nooron.org/know/famous_quotations
http://www.nooron.org/know/internet_ports_and_protocols (very broken)
are from:
http://www.aiagentz.com/brains.jsp (02)
I am not a Brain user myself [that came out wrong] and I would quite
like to get my hands on some Brain files which are rich in associations
since all the AiAgentz examples are essentially just flat databases. If
anybody has Brain files (.brn) that they'd like to see published through
Nooron, drop me a line. It would really help me out, thanks! Likewise
for sitebrains. (03)
<zombie>Brainz... Brainz...</zombie> (04)
--
====================================================================
Shawn Murphy http://www.smurp.com
mailto:smurp@smurp.com http://www.nooron.org
tel:+780-903-2428 http://www.noosphere.org (05)
--
This message is archived at: (06)
http://collab.blueoxen.net/forums/cgi-bin/mesg.cgi?a=tools-yak&i=1052852212.19680.1747.camel@prometheus (07)
|