[ExI] free lisp textbook, was: RE: ai class at stanford

Emlyn emlynoregan at gmail.com
Fri Aug 26 05:21:01 UTC 2011


On 26 August 2011 14:19, Adrian Tymes <atymes at gmail.com> wrote:
> On Thu, Aug 25, 2011 at 9:11 PM, Emlyn <emlynoregan at gmail.com> wrote:
>> http://point7.wordpress.com/2010/11/23/deep-learning-2-0/
>> http://point7.wordpress.com/2011/08/13/follow-up-to-deep-learning-2-0/
>
> This sounds similar to http://www.wikihow.com/ .  I would be interested
> to get your thoughts on comparing your post to what they're going for.

Interesting. wikiHows are similar to google's Knols, no?

These guys are close to the basics of what I'm thinking; they have
related wikiHows. But I think you need dependencies "required" rather
than the weaker "related". With "required", you can plot a path from
what you know to what you want to know; given a set of things you know
(Knows), and a Knows/Learns dependency graph, you can find the set of
shortest paths through the graph from the thing you want to know, back
to things you do know, then list all the "learns" encountered along
those paths in order, and lay it out as a list of things to learn.

wikiHow and Gnols were both just the "Learns" under my nomenclature,
each leading to one "Know" (described by the title). Collapsing
"Learns" and "Knows" together is simpler, but I think you then can't
capture more complex relationships between ways of learning things
(notably, one "learn" might give you multiple "knows", and multiple
"learns" could get you to the same "know".

Did you read the stuff about this bootstrapping based on being a tool
for autodidacts to organise their own learning? I'm really interested
in feedback on this, I think it's a crucial part; it's got to be
useful.

-- 
Emlyn

http://my.syyn.cc - Synchonise Google+, Facebook, WordPress and Google
Buzz posts,
comments and all.
http://point7.wordpress.com - My blog
Find me on Facebook and Buzz




More information about the extropy-chat mailing list