The AI Engine

Thanks, prof.

I just read that stuff about ,eh? getting translated. I am disappointed. I have a keyphrase that tries to notice Canadians. I guess I'll have to get more specific, eh?

Does "huh" turn into "maybe" in preprocessing? If so, why?

I have a keyphrase "I (verb)" with a value of minus 2 and "I" with a value of 1, and several I "verb" with specific verbs, ranging from 2 to 20. I think I read, prof, that a specific key phrase gets preference mostly over the I (verb) knd... And yet, though I keep lessening the value of "I (verb)", when I see the response come up
too often, it still manages to win! How low should I go to differentiate properly between eg I want and I (verb)?

I (verb) is minus 5 now, I is one, and yet... does a two part keyphrase always win over a single maybe?

would it be possible to have seperate xnone keyphrases for different sentance types? like interogative, imperitive etc.?

xnone only comes into play when there is not a single key phrase found... so if you do your job right it goes to a much better "answer" than an xnone would and could ever be.

Yeah, I think that it would be a lot easier to try to catch interrogative questions at the keyphrase level, and not try to get the engine to do it. Most questions come in a set form anyway, especially now with the processing that's done before it gets to our bots.

leetaxx0r- you can delete your plug-ins now. I havent tested it, but it should work.

lunar22- Part of the internal ranking is the length of the Keyphrase. One rank point it added for each letter. So if you want "i want" to supercede "i (verb)" it has to be ranked more than two more. The number in parenthesis in Debug mode by the Keyphrase is the Rank + Length total.

leetaxx0r- You found another one of my plans. I'm going to soon create xcommand which is a special xkeyphrase to deal with commands. It's going to be on a new page, because the response is going to depend on emotional state. For example, if I say "Eat this" to somebody who dislikes me, he/she probably wont, but if I say it to a friend, he/she probably will.

That, or I'll add that ability to all Keyphrases- emotional ranges.

Im new here and I dont know what to do.

emotional ranges for keyphrases would be great. I have an idea for somehting similar but a little different. When someone asks my bot "what's up?" I want it to have plenty of different responses, but if they ask it again, it'll say something different, and thus usually will contradict itself.

Would it be possible to make certain "moods" or something that rotate randomly every certain amount of time and make it so we could have certain responses for each rotation. It would just help make the bots mroe consistent I think. That combined with emotional ranges in keyphrases would make them a lot more realistic I think.

well the "moods" have started to come out already, just look on the My Bots page, with the smiley rating hehe