Virtual Humans Forum
Virtual Humans Forum
Home | Profile | Register | Active Topics | Members | Search | FAQ
Username:
Password:
Save Password
Forgot your Password?

Virtual Humans
Partner website:
Chatbots.org
 All Forums
 Virtual Humans
 Artificial Intelligence
 human mind map
 New Topic  Reply to Topic
 Printer Friendly
Previous Page | Next Page
Author  Topic Next Topic
Page: of 25

laackejim
Committed Member



USA
3274 Posts

Posted - May 20 2008 :  07:12:01  Show Profile  Reply with Quote
Nice to hear from you again Tom. I have been inspired to dig into my own concepts again and have been working toward getting something functional at some level. Have you accomplished anything on that "teacher function" you talked about once? Sorry that my suggestions were a little flat. I will get back around to the concept again, hopefully soon, and try and come up with something more elegantly functional.

Uncle Jim (e=mc2)
Go to Top of Page

toborman
Hooked Member



USA
290 Posts

Posted - May 21 2008 :  03:47:08  Show Profile  Visit toborman's Homepage  Reply with Quote
Harry's original teacher function merely expressed facts he found at random, then formulated questions for the student based on those facts. Not particularly intelligent, but it did allow me to develop the conversation procedure he now uses.

I put aside upgrading the teacher function so I could focus on a reasoning procedure that handles all three types of inference, and can, somewhat easily, learn new inference methods. Test results so far have demonstrated the viabilty of my latest design. (it's neat!)

http://mindmap.iwarp.com
Go to Top of Page

mikmoth
Moderator



USA
2082 Posts

Posted - May 21 2008 :  04:45:01  Show Profile  Visit mikmoth's Homepage  Reply with Quote
Very neat!

Try explaining women to him... if he can figure that out you know he can solve anything!

 http://lhandslide.com
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - May 21 2008 :  08:19:46  Show Profile  Reply with Quote
Hot Dog, progress!

It sounds like I misunderstood your original question or description of the teacher function. You have just triggered some different thoughts in my head. Hmmmmm.

Uncle Jim (e=mc2)
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - May 21 2008 :  08:21:53  Show Profile  Reply with Quote
MIk, women are no stranger than men. They are just in that alternative universe. From their side of the "rift in the time/space continuum" we are just as strange. I hope.

Uncle Jim (e=mc2)

Edited by - laackejim on May 21 2008 08:22:23
Go to Top of Page

toborman
Hooked Member



USA
290 Posts

Posted - May 22 2008 :  03:04:06  Show Profile  Visit toborman's Homepage  Reply with Quote
Harry : hello Tom.
Tom : kari is a human.
Harry : using superset member, I deduced remember kari is a mammal.
Harry : using set member attribute, I deduced remember kari is mean.
Harry : using remove redundant attribute, I deduced forget kari is mean.
Harry : using affirm particular, I deduced remember some humans are mean.
Harry : some humans are mean.
Harry : kari is a mean mean mammal human.
Tom : kari is nice.
Harry : using affirm particular, I deduced remember some humans are nice.
Harry : some humans are nice.
Harry : using deny negative universal, I induced forget humans are not nice.
Harry : using affirm universal, I induced remember humans are nice.
Harry : humans are nice.
Harry : kari is nice.
Tom : kari operates computers.
Harry : I do not understand kari operates computers.
Tom : kari likes movies.
Harry : I do not understand kari likes movies.
Tom : kari.
Harry : I do not understand kari.


You were right. Harry doesn't understand women. I have a lot of work ahead.

http://mindmap.iwarp.com
Go to Top of Page

mikmoth
Moderator



USA
2082 Posts

Posted - May 22 2008 :  15:10:52  Show Profile  Visit mikmoth's Homepage  Reply with Quote
Well hey, I think you're headed in the right direction... with a lot more data I bet it could come pretty close to solving what the heck makes 'em tick. I for one think they are supreme actors... or Jim could be right... they are from another dimension. If I could switch the Higgs field on them they would probably disappear...

 http://lhandslide.com
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - May 24 2008 :  04:30:27  Show Profile  Reply with Quote
Tom (The Toborman)

I have the suspicion that some of the conversation with Harry (5/22) is pulling the collective cyberleg. However, I note superset member, set member, attribute particular, universal, affirm (attribute), deny (attribute). These sure make a lot of sense. How is "some" applied? Is there a "broad" or "general" or "partial" to balance the positive and negative "universal"?

I have just set some code parameters for the "sometimes, always, never" things. In my case I may be too buried in my own predeliction toward numerical encoding to offer any thoughts but Harry's conversation (above) has some very interesting clues in it.

Uncle Jim (e=mc2)
Go to Top of Page

toborman
Hooked Member



USA
290 Posts

Posted - May 24 2008 :  11:06:50  Show Profile  Visit toborman's Homepage  Reply with Quote
quote:
I have the suspicion that some of the conversation with Harry (5/22) is pulling the collective cyberleg.


Although I chose the topic for amusement, Harry’s answers were his own.


quote:
However, I note superset member, set member, attribute particular, universal, affirm (attribute), deny (attribute). These sure make a lot of sense. How is "some" applied? Is there a "broad" or "general" or "partial" to balance the positive and negative "universal"?


Harry is reasoning using propositions. All English inputs are interpreted to propositions and all outputs translated from propositions to English. An English-Proposition translation table guides the process. An inference table contains rules that control reasoning. These rules are deductive, inductive, and abductive in nature and are named as in the example above. All of the propositions are “quantified” with the, all, some, and none, as well as, “certainty” indicators like certain, probable and possible. Some of the rules have been derived from Aristotle’s Square of Opposition that uses the terms all, some and none, and universal and particular. The Square of Opposition provides the rules for deductive inference on categorical propositions, using these terms.

I’m planning to add numerical quantifiers for some. These will be a ratio of the sum of positive and negative examples. These will be used to predict the “likelihood” of future examples.


quote:
I have just set some code parameters for the "sometimes, always, never" things. In my case I may be too buried in my own predeliction toward numerical encoding to offer any thoughts but Harry's conversation (above) has some very interesting clues in it.


It looks like you are working with temporal events. I haven’t started the task of representing present, past, future and conditional. I think I’ll be using the ratio approach and the Square of Opposition for reasoning about time-based events.

http://mindmap.iwarp.com
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - May 25 2008 :  05:32:35  Show Profile  Reply with Quote
I simply draw the temporal inference from the input phrasing. Nothing more than past, present, future past (my way of accommodating "I will have been" etc) and future. This future breakdown has some inherent probability assumptions.

It also significantly limits the topic, action, and tense options for any Present condition.

I am ashamedly ignorant of Aristotle's Square of Opposition but the terms all, some, none are perhaps too obvious because they sound simple and straight forward. Universal and particular would seem the same except for the obvious overlap with all, some and none. Some seems now to be the odd man out. Well worth some thinking and learning on my part.

I my case the "logic" of induction, deduction, and abduction (something tells me I don't understand that last one) is context driven during communication. The context is accumulated from the person, tense, action, activity and emotion of the conversation topic. It is that context that provides the flags to identify concepts, responses, and information as being anomalous or "illogical". Probably because of my own brain I have not been able to find a stable place for classical rules of logic. It seems that most communication, and the reasoning that goes behind it, falls within the "if it is consistent it is probable, if it is odd it is possible, if it is inconsistent it is improbable".

I would love to watch Harry grow -- around women -- with humor.


Uncle Jim (e=mc2)
Go to Top of Page

toborman
Hooked Member



USA
290 Posts

Posted - May 25 2008 :  10:17:04  Show Profile  Visit toborman's Homepage  Reply with Quote
Deductive inferences are certain. Inductive inferences are probable. Abductive inferences are possible.

Universal affirmative = all. Universal negative = none. Particular affirmative= some are. Particular negative = some are not. Some is interpreted as at least one.

http://mindmap.iwarp.com
Go to Top of Page

laackejim
Committed Member



USA
3274 Posts

Posted - May 25 2008 :  20:54:50  Show Profile  Reply with Quote
Got it. Thanks Tom. That gives me a little jolt. In the flow design for my concept it appears that I have an inherent error. I am not sure where it will turn up but the cleanness of the definitions you gave make me think I am going to have a wrinkle I wonder if that mental quiver is coming from the conflict between clear logic and fuzzy (read "possibilistic") reasoning.

You always jar my thinking. It is always valuable. Most of the time it is fun.

Uncle Jim (e=mc2)

Edited by - laackejim on May 25 2008 20:55:50
Go to Top of Page

toborman
Hooked Member



USA
290 Posts

Posted - Jul 31 2008 :  04:38:38  Show Profile  Visit toborman's Homepage  Reply with Quote
For those of you working on AI engines, here is a starting set of deductive processes.
P, q, r, and s are propositions. > means implies, - means not, v means or, & means and.
To use these methods you must convert your sentences into propositions.
These methods can be used to create or validate arguments, or infer propositions.

Name... Method
Modus Ponens... p > q, p; therefore q
Modus Tolens... p > q, -q; therefore p
Chain... p > q, q > r; therefore p > r
Disjunctive1... p v q, p; therefore q
Disjunctive2... p v q, p; therefore q
Addition1... p; therefore p v q
Addition2... q; therefore p v q
Conjunctive1... -(p & q), p; therefore -q
Conjunctive2v -(p & q), q; therefore -p
Simplification1... (p & q); therefore p
Simplification2... (p & q); therefore q
Adjunction... p, q; therefore p & q
Reductio1... p > -p; therefore -p
Reductio2... p > (q & -q); therefore -p
Complex constructive... p > q, r > s, p v r; therefore q v s
Complex destructive... p > q, r > s, -q v -s; therefore -p v -r
Simple constructive... p > q, r > q, p v r; therefore q
Simple destructive... p > q, p > r, -q v -r; therefore -p

I'll post inductive and abductive methods later.

http://mindmap.iwarp.com
Go to Top of Page

GrantNZ
Dedicated Member



New Zealand
2677 Posts

Posted - Jul 31 2008 :  12:39:07  Show Profile  Reply with Quote
Thank you toborman, that's useful stuff
Go to Top of Page

toborman
Hooked Member



USA
290 Posts

Posted - Jul 31 2008 :  19:14:21  Show Profile  Visit toborman's Homepage
Here is an example that shows how to interpret the methods described above.

Modus ponens… p>q, p; therefore q

p>q… If rain falls on street then the street is wet.
p… Rain falls on the street.
Therefore
q… The street is wet.

http://mindmap.iwarp.com
Go to Top of Page
Page: of 25  Topic Next Topic  
Previous Page | Next Page
 New Topic  Reply to Topic
 Printer Friendly
Jump To:
Virtual Humans Forum © V.R.Consulting Go To Top Of Page
This page was generated in 0.15 seconds. Snitz Forums 2000