heartlogic-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Heartlogic-dev] interesting developments


From: William L. Jarrold
Subject: [Heartlogic-dev] interesting developments
Date: Sat, 21 May 2005 01:46:07 -0500 (CDT)

Hi Guys,

Had an interetsing chat with someone from Cycorp earlier today. Some positive developments.

1) it is pretty easy to generate english assertions from cycl assertions and it works pretty well. the cycorp guy generated about 20 at random
while we were talking and read them to me.  all we very comprehensible.
about 70 % of them were about terrorists so we can find away to filter out
those assertions. bottom line: at least some aspects of operatiom rumination pilot will scale nicely...this means we can probably easily
get up to 100 assertions or more to test without even getting research
cyc.

joshua, do you think you can find a way to write a program that will
return a random assertion from the open mind...if we can make this
work, we can easily do some interesting comparitive studies.

2) the legal issues of getting research cyc and putting engilsh random
assertions on the web for random people to judge should not be
insurmountable as far as cycorps legal/financial interests are
concerned.  they just don't want any jamoke to download the entire cyc
for free w/o at least signing some kind of agreement.  but since the
random folks visiting our website will only see a few assertions at a
time, this will be fine.

3) this person had an idea that made me think of this: don't give them
rules to rae.  rather look at conclusions and decide what is better --
fix the rule or fix the hags that support the ruels.

bill




reply via email to

[Prev in Thread] Current Thread [Next in Thread]