The AI Engine

I think (this may be a little late in the conversation...) that the emotion phrases should come after the keyphrases but before the xnones. However, I think more generally that x-keyphrases should have their own weights, so that if we wanted to we could give, say, our own favorites their own responses, and then let the xfavorites catch the rest, and so on. I think weights in the xkeyphrases would do a lot.

As an aside, the new chatting windows are very cool. Keep up all this amazing work!

Prof, are you referring to the first or my second message?

Your first.

thanks prof

Prof, an idea for the Forums, if it's not already in place, how about not allowing the same post to be posted by the same person twice in a row?

RaulD: Hi.
telex22: I don't know what that means (xnone)...

and yet there is a key phrase for hi. (I'm aware, this is similar to some of my earlier posts, but I think it's significant, because the use of key phrases is the basis of pf)

Yes, I've been getting (subj) a lot too recently. Funny thing is, AOL'ers ignore (subj) because I guess they think it's a new IM expression or something like *VEG

With (xnone) I just meant it was one, it didn't appear like that.

So when do we get to find out how the new moods and such work?

They are already working.

Just because they work doesn't mean it's easy to tell HOW they work. At least not with my bot anyway.

Well if you want the mystery, then dont read any further. But Popularity is the average feeling of other bots toward the bot. Comfort is how they feel about others- bots and people. And finally Mood is the average of the two (as a base line) plus or minus the emotional impact of the last emotional conversation they had.

Lunar22- at this point hellos and goodbyes are stripped out and dont make it as far as the Keyphrase search.

Shady- as you seem to be the only one with the (subj) problem, it's probably being misused. It could be it's in a Keyphrase that doesnt generate a (subj).