Date:

Share:

Replace Your Friends with a Markov Chain Slackbot

Related Articles

TL; DR https://github.com/perciplex/new-friend This is a slackbot that can be trained on your loose history to emulate your friends and co-workers for great fun for everyone.

You should check out Ben’s blog post which is much more in depth with beautiful diagrams http://blog.benwiener.com/programming/2021/07/10/new-friend.html See?

My friends and I have an LLC side event called Plexiglas That we have had for several years. We spent a lot of time working on a project called RaaS (Intermittently for more than a year), which really passed like a lead balloon. I need to write writing about the interesting parts of it.

Then, we wanted to get into easier projects. So here’s our first. We had slackness to our group of friends long enough for each person’s word count to be 100,000 words count. We thought it would be fun for a weird robot channel to show up one day.

Originally we planned to take a stab at using repetitive neural networks. We got up and did it and while we were at it we tried As a collaboration. It worked well enough for it to memorize sentences.

But then we took a step back to try the simpler approach, a simple model of the Markov necklace. We used this library https://github.com/jsvine/markovify That if you look at the content of it is not even that complicated. Basically, you collect a count of how many times a word follows its previous words. You can quite easily build such a thing yourself. https://sookocheff.com/post/nlp/ngram-modeling-with-markov-chains/

To trick a Markov chain into mimicking a conversation, we put together the actual word and the user who spoke it as a “pseudo-word.” We also added a special “to send“Tokens. You can see it here https://github.com/perciplex/new-friend/blob/f73dba011bc589c0ba825b92a4e6460fa28e0f26/new_friend/bot.py#L85

good things.

Source

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Popular Articles