Manual:Personality

From VerbotWiki

Personality Chatbots

Introduction

For creating a personality knowledge base, you should believe, understand, what you are. ,how you mingle with other humans, what makes & broke a relationship etc. some knowledge in psycology is good, but not at all important, because you in your current situation (without knowledge of psycology) is almost feeded with infos on psycology(the difference is that you are not able to distinguish between it), since someone has put a name(psycology) into that particular subject people thinks that they need to study that, it's not like that, you know what you are, but you don't get time to think about that you are busy with your life (Human Psycology has also got so much advanced studies, like cognitive science, mind, dream influences, learining, consciousness and the unconscious etcc) but lets leave the advanced thing there,) if someone ever put these subjects wisely in a KB, then we are sure that, we can see Artificial Intellgence

How to Create?

The best way to create a personality knowlede base is to use your conversational skills, begin by making simple rules like, hi, how are you , what is your name etc. See this is the first or basic conversation which we humans do when we meet someone(a friend, a relative, a stranger etc.) and then proceed with more and more conversational subject, also make it your own personality,(without feeding any personal details)

Motivation

You can not build one, if your aim is to enjoy or test verbot, there should be a motivation, like "i can do this", "i know how to give intelligence to a program", "i understands a what is called human beings, how that thing work etc) (no need to know programming languages) just use your conscious brain , the aim should be to share your creation with others(with this world), otherwise if you are just creating a personality KB for your own use, then you will never finish it)

We will try to bring more infos about this in the future, or You can view This TOPIC (http://www.verbots.com/forums/viewtopic.php?t=831)

__________________________ BlueSpace Argh!... Culture always culture...Innate/inborn or acquired knowledge...

An old story. In opposition to a real brain, 

a machine is a blank/empty page, we can't build an A.I without knowledge of psycology and to apply this to a predef network. Perception and hierarchy is a most importante thing for a machine. It's the raison why, i dream Verbot will be able to count Variables and give many values to a variable (like a cel in Excel). It only can perceive what we type. Not Webcam, no ears. Some basic behaviors. Fear===> avoidance, aggressiveness,lies, withdrawn attitude, fast and speak sternly language, bad learning mode etc etc Joy ===> love, slow and nice language, give, receive, kindness, good learning mode etc etc A chance to build that is to count the variables and give them differents values . Ex: 3 words (key word with a value) about violence like: war(value +20), gun(val +10), mafia (value +15). A word like love (value -25)

Verbot recognize these words and their value. Up to, ex: 30 Verbot can see we are not joking, and it is ready to wear his armour and take the corresponding mood. (buy nuclear-submarine, rockets, tanks, lightnings like Zeus :0)) It's a kind of hierarchy or predef behavior.

I stop here because, it's a very vaste subject.( and i can speak like that during hours and hours.). We have not the tools to do that, may be with C# but i don't know it and no time to learn, my real job is hard enough.

Friendly Sylvain

__________________________

Others ideas We need a moteur/engine easy to use (Ok, not to build)for users of the community, where, like for a hive, everybody can work like bees. Just one thing to do for then, reply to questions, and that, only from the Verbot Player Standard main windows. Without this, we are.... nada. (To add/merge with my post Sun Aug 21, 2005 12:38 pm in "Discussion".)

We have to give value to some words. In French we use about 250 words for commun conversation, but it's a hard and warped language, more easy/easier in English, a man from New-Castle say to me about 150 words. That's means 150 values we have to add to 150 variables [words]. Verbot have only to give 1 value/mark/sign (actually, Aaron said it's not possible to attach 2 values/functions etc for a variable, strange, why this kind of restriction?... But ok, may be later) to each variable. Ex: Actually i try in my automatic learning mode (2 kinds, a "Quick" and a "Deep"(an other .VKB)). For "Quick" learning mode. Conversation: -Me: Zebra -Verbot: i don't know what a Zebra is, do you wish give me infos about this strange Zebra? -Me: Yes (Verbot sends learning mode) -Verbot: Ok, now i'm in learning mode. Is it an animal, a place, a thing or else? -Me: an animal -Verbot: Ok thanks, i note Zebra is animal and i give value AA_1="anim" to zebra. Is it hairy or scaly? -Me: hairy -Verbot: Ok, now i know Zebra is hairy. I give value AA_2="anim"+"hairy" to my data.Does it run fast? -Me: Yes -Verbot: Great! I give value AA3="anim"+"hair"+"rapid" Does it lives on Earth? Etc And that, only because we can't add no more than one value for a variable. Etc /* May be later, when we'll be able to have namy values for a variable. Ex: AA=(animal)_0(first var)_0(second)_0(third) etc

"AA_0_0_0_FuncX(from 0 to 1O)=0" for all animals/ "AA_1_0_0_FuncX(from 0 to 1O)=0" for zebra/ "AA_1_1_0_FuncX(from 0 to 1O)=0" for zebra+hairy/ "AA_1_1_1_FuncX(from 0 to 1O)=0" for Zebra+ hairy+fast/ "AA_0_0_0_FuncX(from 0 to 1O)=0" for all animals/

Etc. Where "F" is an incremental function who counts the importance of an infos. It more important to know a Zebra is fast/rapid than hairy for his survival. If the value is "10" it's a very important info. To "O" we don't care. We can add all the infos we need. Verbot stores stores and stores again.

Later in a conversation, we will be able to find all the "hairy+animal". Ex: -Me: i need infos about "hairy animal". Verbot must replies all infos it have about "hairies+ Animals". It must find all the lines. "AA_1_1_0_FuncX(from 0 to 1O)=0" for animal/hairy

AA_0=All the Animal of the galaxie AA_0_1=hairy AA_0_0_1=fast

And the same for Human. Ex: BB_0=human BB_0_1=hairy BB_0_0_1=fast Etc etc (Later, we will be able to create, from Verbot main windows (Verbot standar Player), variables and give then a several values. In fact it's the final goal)

For "Deep" learning mode, it's not really the same thing... because: date/wood?/steel?/plastis?/job/name/gender/ coulours/gender/place/where/when/how/How many/ etc etc We can't have a line like that: "AA_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_0_ 0_0_0_0_0_0_0_0_0_0_0_0_FuncX(from 0 to 1O)=0" I have other ideas I only say: "Comparison/union/difference/soustraction/ /lower to/upper to/errors/ etc. Ok i stop, Thanks to be patient. At that moment we can speak about a kind of start for psychology behaviour, and it will be a very hard performance, not like this story with, "blabla AA_0_0_0 blabla" who seems easy to build. An example? Thanks! Great! (Thanks for your support) If Verbot find: "AA_1_1_0_0 etc", it's an animal/hairy ok Of course it can't be a reptile, a thing, a human, a planéte, a car, a washing machine, etc... Stop ok ok. I can speak about this subject no stop during 1 month, Verbot has me in its grip.

• /

With that, be shure, Verbot leaves all other Chatter bot far behind. The users have only to answer to Verbot questions after choosed "Quick" mode or "Deep" mode. Easy, no?

Of course i have many others ideas but, at first an excellent moteur/engine, after for general Verbot psychology, it's relatively easy and funny, it's a thing for imagination, not a story with this mysterious C#. Friendly Sylvain