Like Ra in latex catsuit, latex mask and high heels
Like Ra's Naughty Playground

sexy oil shiny tights pantyhose bodysuit
Crotchless Oil Shiny Tube Top Tights Sexy Underwear Bodysuit Smooth Fullbody Open Crotch Pantyhose Medias 1D Bodystockings
$60.51

"leather" hood "bondage"
Strictly Bound #1 Hoodie Long Sleeve Bondage Fetish Bdsm Leather Strict Bondage Erotic Ball Gag Gag Trainer High Heels
$130.08-5%

"Yimunancy" lingerie
Yimunancy Transparent Floral Night Dress Women Sexy Lingerie Set 2-Piece Sexy Erotic Mesh Dress Set
$17.27-40%

"ladyboy"
I love ladyboys in Thai T-Shirt street wear quick drying anime stuff cotton man t-shirts clothes for men
$86.62

"crossdress"
CD Silicone Adhesive Roll-On Emulsion Binder Chest Stickers Invisible Fake Silicone Breast Glue Water Crossdresser Skin Special
$24.18-19%

shiny glossy swimsuit bodysuit onepiece
AMORESY-High-Leg Back-Zip Swimsuit Leotard, Shiny Glossy Tights, One Piece Swimwear
$286.76-10%

shiny catsuit sexy unitard bodysuit
Sexy Women Faux Leather Catsuit PVC Latex Body suit Shini Metallic Jumpsuit Tight Zentai Suit Halloween Cosplay Suit L
$54.58-18%



To view Aliexpress you might need to switch your mobile browser to the Desktop version.


Magnetic lock for self-bondage
Magnetic lock for self-bondage
€145

If you would like to use search functions, view hidden archives or play games, please consider registering.


Artificial Intelligence chatterbot and NLP (Natural Language Processing)
#51
(25 Jan 2015, 15:54 )Like Ra Wrote: the "original" character is very sarcastic. I'm not changing much of the original personality of the bot, but mostly add new knowledge and dialogues.

A trait like that solves a depth problem in the interaction. It's a sort of cheat to allow a more shallow response to a question of greater depth.
I found that when using "microsoft agents" ... If you gave the character more attitude it made there limited responsive depth some how more tolerable.
Actually it would be funny to couple program O to a MS agent... they are web enabled and float over IE ... I can just imagine "Peedy" the green parrot giving bondage advice.
Reply
#52
I need to create proper AIML files and make them work first 😁

As for Program O - I logged 3 bugs, 2 are aeady resolved. However <topic> still does not work.
Reply
#53
Just learned that AIML interpreters (aka AIML ChatBots) are used to write stories. This is exactly what's needed for automatic self-bondage scenario/session generator.

Let's collect some self-bondage ideas here: Self-bondage session generator
Reply
#54
Just in case if you are curious - yes, I'm working on the chatbot, constantly updating "the brain". Also, thanks for testing - I'm using the dialogues to enhance the responses and add more capabilities.

Meanwhile I logged more bugs for Program O.

As always, your ideas are welcome.
Reply
#55
The good thing is that I can update my CV with AIML knowledge, I think I tested and played with at least 90% of it and have read about the rest 😁

The bad thing is that there are even more bugs and also some standard features (AIML 1.0.1) are not supported. AIML v2 is even more fun. But Program O does not work with v2.

Let's see what the developers say...
Reply
#56
Guys, thanks for some beautiful bondage dialogues 😉 Now I have some data (and bugs) to tweak naughty Miyuki's brain again!

Now I have to work in two more directions - the dominant and the submissive sides, not only "reflective" and "advice giving" ones.
Reply
#57
I went through all latest chat logs and added/fixed/changed as much as I could. Currently I'm trying to make Miyuki asking the same questions. The only possibility to achieve that is to set/check variables. And considering there are bugs in this area, it's not that easy.

BTW, as a test you can ask: "Tell me what I like"
Reply
#58
Thanks much Marcus!!!

Quote:Miyuki: Any particular kind of leotards and pantyhose?

ROTFL!!!!!

One your conversation has given me a couple of days of hard work 😁

Yes, clearing Miyuki's mind process is in progress, but not complete yet. I'm thinking how to properly reset the variables. And yes, I'm adding more variations. BTW, there are almost 100,000 patterns and corresponding responses defined, so it's not an easy task. But fun, no doubts 😉

Just a note. Usually bots cannot follow two different responses in one phrase. Too difficult to implement properly.
Reply
#59
Oh yes, and spelling must be correct, though some spelling corrections are implemented.
Reply
#60
And those pesky Program O bugs....
Reply




Contributors: brandynette (12) , cinon (1) , dhf7b8g (4) , egregious (1) , Highinheels66 (2) , HypnoMix (3) , Lancer (8) , Like Ra (203) , lugnuts (1) , madjack (12) , Marcus (25) , mcswitchy (2) , MstrChain (1) , napras-sword (1) , shinybambi (2) , spawn (9) , Tinker D (14) , Zooy (2)