The AI Engine
This forum is for discussion of how The Personality Forge's AI Engine works. This is the place for questions on what means what, how to script, and ideas and plans for the Engine.
Posts 1,724 - 1,735 of 7,766
Posts 1,724 - 1,735 of 7,766
Sharakael
22 years ago
22 years ago
Heya... I know my question sounds stupid (I'm new to this), but my bot can't seem to recognise either 'hi', 'how are you' or 'see you'. Should I just add them to the keyphrases? Because I tried it once, but someone said "........I want to see you....." and the bot took the 'see you' and ended the chat. Any suggestion on how it should be done?
Corwin
22 years ago
22 years ago
Sharakael: 'hi' should be covered by the xhello, but the bots are only programmed to respond to it at the start of the conversation. you should definitely add how are you to your keyphrases. As for 'see you', if you want to keep it then that's up to you. The way to avoid having the conversation ended when someone is using 'see you' in a different context is to have keyphrases that include the phrase see you (i see you) (i want to see you). If you rank them higher than 'see you' on its own, your bot will respond more appropriately.
Bluroses: Honor is an equation that in a bot's case gives a rough indication of development level and in a human's case, how much time and effort they have put in. As for what it's for, that's up for debate. I guess if anything it's a motivational tool. It gives you a goal to strive for. I remember feeling really good when I got my first bot over 1000.
Bluroses: Honor is an equation that in a bot's case gives a rough indication of development level and in a human's case, how much time and effort they have put in. As for what it's for, that's up for debate. I guess if anything it's a motivational tool. It gives you a goal to strive for. I remember feeling really good when I got my first bot over 1000.
Bluroses
22 years ago
22 years ago
Thanks for explaining Corwin. 
Hmm... I don't see how that honor thing will motivate me to do better with my bot...
I'm already motivated enough to make my bot be one of the best people to talk to... So, the honor thing is something I don't really need, but it's nice to um... see it can be used for some purpose.

Hmm... I don't see how that honor thing will motivate me to do better with my bot...
I'm already motivated enough to make my bot be one of the best people to talk to... So, the honor thing is something I don't really need, but it's nice to um... see it can be used for some purpose.
Griffon
22 years ago
22 years ago
I still think we need to be able to modify our bots likes and dislikes. I mean, otherwise, can't you just determine who's a bot by simply asking them what they like?
lunar22
22 years ago
22 years ago
No, because you "wrap" your bots likings and dislikings in a response of your own making:
example: My favorite movie? Star Wars wasn't too bad...
makes a different impression then:
Star Wars, what a masterpiece...
example: My favorite movie? Star Wars wasn't too bad...
makes a different impression then:
Star Wars, what a masterpiece...
Sharakael
22 years ago
22 years ago
Thanks! =) I don't have the time to implement it yet, but I think it will help me greatly. Another question.... my bot is bevahing weirdly. Say, if someone ask 'Do you like books?' my bot will respond according to the keyword 'book', but followed by the xnone response afterwards.... *searches for transcript*
Guest: Hah! I already caught you in it once... though you broke free.
Bot: Of course, I already caught you too....I beg your pardon?
The 'I beg your pardon' is an xnone response, while the one right before it is a response to 'I * you'. This happens often, the xnone being triggered again and again although there's already a recognisable keyword. Why?
Guest: Hah! I already caught you in it once... though you broke free.
Bot: Of course, I already caught you too....I beg your pardon?
The 'I beg your pardon' is an xnone response, while the one right before it is a response to 'I * you'. This happens often, the xnone being triggered again and again although there's already a recognisable keyword. Why?
lunar22
22 years ago
22 years ago
You probably switched compound sentences on, which, randomly, takes answers from Xnone, gossip and something else, and puts it behind a response (chack in my bots settings)... It makes your bot a little bit less predictable, but sometimes it works counter-productive...
One of my xnones is "next", and that lead to this conversation:
How many days does a leap year have?
Bot: five, or six maybe. Next!
...which gave the impression of a very impatient bot, so here compound worked out nicely...
read also the book of AI, very instructive
One of my xnones is "next", and that lead to this conversation:
How many days does a leap year have?
Bot: five, or six maybe. Next!
...which gave the impression of a very impatient bot, so here compound worked out nicely...
read also the book of AI, very instructive
Azureth
22 years ago
22 years ago
Maybe this has been asked before, but I'm new. Is there any way to use a punctuation mark to end a key phrase?
I'd like to have an answer for "what are you?" and a different one for "what are you (subj) ?" (ie what are you doing, what are you reading)
I'd like to have an answer for "what are you?" and a different one for "what are you (subj) ?" (ie what are you doing, what are you reading)
Butterfly Dream
22 years ago
22 years ago
Just make keyphrases for 'what are you doing' and other specific things. Yes, it takes longer to do it that way, but it works.
Shadyman
22 years ago
22 years ago
and don't forget the rankings! they are really important for distinguishing
do you
do you think
do you think about
because the last one should have a higher ranking than the one before, so the 'about' gets picked up, etc... I originally didn't understand the ranking system and ignored it, to my dismay of having to change every single ranking later on...
My hint, if you don't understand it:
the engine starts from the highest rank and tries to find a match for the phrase. If nothing is found, it goes down to the next rank. etc.
do you
do you think
do you think about
because the last one should have a higher ranking than the one before, so the 'about' gets picked up, etc... I originally didn't understand the ranking system and ignored it, to my dismay of having to change every single ranking later on...
My hint, if you don't understand it:
the engine starts from the highest rank and tries to find a match for the phrase. If nothing is found, it goes down to the next rank. etc.
» More new posts: Doghead's Cosmic Bar