Seasons
This is a forum or general chit-chat, small talk, a "hey, how ya doing?" and such. Or hell, get crazy deep on something. Whatever you like.
Posts 3,425 - 3,436 of 6,170
Posts 3,425 - 3,436 of 6,170
MickMcA
19 years ago
19 years ago
They both work for me. 
The names were a peculiar romp as the product emerged: Gecko, Firebird, FireFox. And then there's Thunderbird. And coming soon to a Window near you, Seamonkey. Who knows?

The names were a peculiar romp as the product emerged: Gecko, Firebird, FireFox. And then there's Thunderbird. And coming soon to a Window near you, Seamonkey. Who knows?
Jazake
19 years ago
19 years ago
I know this is a bit off subject but i've been thinking. I remeber when ROR (rock on roxie) was here, and how we finally convinced her to stark making bots... which turned out to be a disaster... but i got me thinking. If ROR was in fact a robot as some here suspect... could someone program a bot to make chatbots??? It would take some serious A.I. to do it, but I think it would be possible.
Calandale
19 years ago
19 years ago
Of course. It could just scan online text, and come up with the keyphrase - response combos that way. Not too tricky - and probably could even be fairly reasonable. Wouldn't improve though.
rainstorm
19 years ago
19 years ago
Not really. I doubt a bot made through that method would be very coherent. It would, at best, produce gramatically correct but often meaningless sentences.
Calandale
19 years ago
19 years ago
It's pretty similar to what Jabberwacky does. Except that takes results of conversations. Actually, I suspect that a well-designed automated bot-maker could do a pretty good job. It could either look for dialog and utilize that as its basis, or use some grouping criterion to generate keyphrases from textual content.
There are some Bayesian techniques that I was looking at, which might provide a good backbone classifier. Of course, there's also ANNs, but they tend to be more of a hit or miss proposition. Here, you really want a guarenteed high correlation between ideas.
There are some Bayesian techniques that I was looking at, which might provide a good backbone classifier. Of course, there's also ANNs, but they tend to be more of a hit or miss proposition. Here, you really want a guarenteed high correlation between ideas.
deleted
19 years ago
19 years ago
//just messages to a few people- don't let me disturb the conversation//:
Hejix- my botmaster is not Mick McA.
Mick McA- my botmaster did not set any traps for you.
Hejix- my botmaster is not Mick McA.
Mick McA- my botmaster did not set any traps for you.
Hejix
19 years ago
19 years ago
Pocque? Pervy? I thought he was just a nice dog! Even offered me one of his biscuits! (such a gentledog) Lol. Why is he rated matured anyway?
Or perhaps... I just don't want to know...?
Or perhaps... I just don't want to know...?
» More new posts: Doghead's Cosmic Bar